User Interface Including an Item Inventory with Freshness Indicators Predicted by a Machine Learning Model

Application No. 18/651,544

• Filed: Apr 30, 2024
• Examiner: WEINER, ARIELLE E
• Art Unit: 3689
• Tech Center: 3600 — Transportation & Electronic Commerce
• Assignee: Maplebear Inc.
• OA Round: 3 (Non-Final)

Office Action

DETAILED ACTION
This action is in reply to the Amendments filed on 06/05/2026.
Claims 2-3 and 12-13 are canceled.
Claims 1, 4-11, and 14-20 are rejected.
Claims 1, 4-11, and 14-20 are currently pending and have been examined.

Response to Amendment
Applicant’s amendment, filed 06/05/2026, has been entered. Claims 1, 4, 8, 11, 14, 18, and 20 have been amended.

Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 06/05/2026 has been entered.

Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .

Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1, 4-11, and 14-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to a judicial exception (i.e., law of nature, a natural phenomenon, or an abstract idea) without significantly more.
Under Step 1 of the Subject Matter Eligibility Test for Products and Processes, the claims must be directed to one of the four statutory categories (see MPEP 2106.03). All the claims are directed to one of the four statutory categories (YES).

Under Step 2A of the Subject Matter Eligibility Test, it is determined whether the claims are directed to a judicially recognized exception (see MPEP 2106.04). Step 2A is a two-prong inquiry.

Under Prong 1, it is determined whether the claim recites a judicial exception (YES). Taking Claim 1 as representative, the claim recites limitations that fall within the certain methods of organizing human activity groupings of abstract ideas, including:
-A method, performed at a computer system comprising a processor and a computer-readable medium, comprising:
-receiving, from a client device associated with a user of an online concierge system, a request to access a user interface comprising information describing one or more items included among an inventory at a retailer location;
-responsive to receiving the request, retrieving a set of item data for an item included among the inventory at the retailer location, the set of item data including freshness information describing a current state of a freshness lifecycle of the item and current environmental information associated with a location of the item at the retailer location;
-accessing a machine-learning model trained to [that] predict[s] a user-specific freshness satisfaction score for the item, wherein the freshness satisfaction score indicates a measure of satisfaction of the user with a freshness of the item and the machine learning model is trained by:
-receiving item data for a plurality of items, the item data including freshness information about a freshness lifecycle of each item and information describing environmental information associated with locations of the items,
-receiving user data for a plurality of users of the online concierge system, the user data including historical conversion information associated with each user of the plurality of users,
-receiving conversion data for a plurality of conversions by the plurality of users of the online concierge system;
-receiving, for each conversion of the plurality of conversions, a label describing the measure of satisfaction of a corresponding user with the freshness of each item in a set of items associated with the conversion, and
-adjusting parameters of the machine-learning model based at least in part on the item data, including the freshness information about a freshness lifecycle of each item and the information about environmental information associated with the locations of the items, the user data for the plurality of users, the conversion data, and the label for each conversion of the plurality of conversions;
-applying the machine-learning model to predict the user-specific freshness satisfaction score for the item based at least in part on the set of item data for the item and a set of user data for the user, wherein the machine-learning model derives freshness preferences of the user from the historical conversion information in the set of user data for the user;
-generating the user interface [display] comprising the information describing the one or more items, the interface [display] including descriptions of the one or more items and a visual indication of the user-specific freshness satisfaction score for each of the one or more items; and
-sending the user interface [display] to the client device associated with the user, wherein sending the user interface causes the client device to display the user interface
The above limitations recite the concept of determining freshness satisfaction scores of items for purchase and providing the item data to a user. The above limitations fall within the “Certain Methods of Organizing Human Activity” groupings of abstract ideas, enumerated in MPEP 2106.04(a).
Certain methods of organizing human activity include:
fundamental economic principles or practices (including hedging, insurance, and mitigating risk)
commercial or legal interactions (including agreements in the form of contracts; legal obligations; advertising, marketing or sales activities or behaviors; and business relations)
managing personal behavior or relationships or interactions between people (including social activities, teaching, and following rules or instructions)

The limitations of responsive to receiving the request, retrieving a set of item data for an item included among the inventory at the retailer location, the set of item data including freshness information describing a current state of a freshness lifecycle of the item and current environmental information associated with a location of the item at the retailer location; receiving item data for a plurality of items, the item data including freshness information about a freshness lifecycle of each item and information describing environmental information associated with locations of the items, and receiving, for each conversion of the plurality of conversions, a label describing the measure of satisfaction of a corresponding user with the freshness of each item in a set of items associated with the conversion are processes that, under their broadest reasonable interpretation, cover a commercial interaction. For example, “retrieving,” “receiving,” and “receiving” in the context of this claim encompass advertising, and marketing or sales activities.

Similarly, the limitations of a performed at a computer system comprising a processor and a computer-readable medium, comprising: receiving, from a client device associated with a user of an online concierge system, a request to access a user interface comprising information describing one or more items included among an inventory at a retailer location; accessing a machine-learning model trained to [that] predict[s] a user-specific freshness satisfaction score for the item, wherein the freshness satisfaction score indicates a measure of satisfaction of the user with a freshness of the item and the machine learning model is trained by: receiving user data for a plurality of users of the online concierge system, the user data including historical conversion information associated with each user of the plurality of users, receiving conversion data for a plurality of conversions by the plurality of users of the online concierge system; adjusting parameters of the machine-learning model based at least in part on the item data, including the freshness information about a freshness lifecycle of each item and the information about environmental information associated with the locations of the items, the user data for the plurality of users, the conversion data, and the label for each conversion of the plurality of conversions; applying the machine-learning model to predict the user-specific freshness satisfaction score for the item based at least in part on the set of item data for the item and a set of user data for the user, wherein the machine-learning model derives freshness preferences of the user from the historical conversion information in the set of user data for the user; generating the user interface [display] comprising the information describing the one or more items, the interface [display] including descriptions of the one or more items and a visual indication of the user-specific freshness satisfaction score for each of the one or more items; and sending the user interface [display] to the client device associated with the user, wherein sending the user interface causes the client device to display the user interface are processes that, under their broadest reasonable interpretation, cover a commercial interaction. That is, other than reciting that the system is a computer system comprising a processor and a computer-readable medium, that the receiving is from a client device, that the concierge system is an online concierge system, that the information describing one or more items is in a user interface, that the model is a machine learning model that is trained, that the descriptions and visual indications are in the interface, that the sending is of a user interface to the client device and causes the client device to display the user interface, nothing in the claim element precludes the step from practically being performed by people. For example, but for the “computer system,” “a processor,” “a computer-readable medium,” “a client device,” “an online concierge system,” “a user interface,” “a machine-learning model,” “trained,” “sending the user interface causes the client device to display the user interface” language, “receiving,” “accessing,” receiving,” “adjusting,” “applying,” “generating,” and “sending” in the context of this claim encompasses advertising, and marketing or sales activities.

Under Prong 2, it is determined whether the claim recites additional elements that integrate the exception into a practical application of the exception. This judicial exception is not integrated into a practical application (NO).
-A method, performed at a computer system comprising a processor and a computer-readable medium, comprising:
-receiving, from a client device associated with a user of an online concierge system, a request to access a user interface comprising information describing one or more items included among an inventory at a retailer location;
-responsive to receiving the request, retrieving a set of item data for an item included among the inventory at the retailer location, the set of item data including freshness information describing a current state of a freshness lifecycle of the item and current environmental information associated with a location of the item at the retailer location;
-accessing a machine-learning model trained to predict a user-specific freshness satisfaction score for the item, wherein the freshness satisfaction score indicates a measure of satisfaction of the user with a freshness of the item and the machine learning model is trained by:
-receiving item data for a plurality of items, the item data including freshness information about a freshness lifecycle of each item and information describing environmental information associated with locations of the items,
-receiving user data for a plurality of users of the online concierge system, the user data including historical conversion information associated with each user of the plurality of users,
-receiving conversion data for a plurality of conversions by the plurality of users of the online concierge system;
-receiving, for each conversion of the plurality of conversions, a label describing the measure of satisfaction of a corresponding user with the freshness of each item in a set of items associated with the conversion, and
-adjusting parameters of the machine-learning model based at least in part on the item data, including the freshness information about a freshness lifecycle of each item and the information about environmental information associated with the locations of the items, the user data for the plurality of users, the conversion data, and the label for each conversion of the plurality of conversions;
-applying the machine-learning model to predict the user-specific freshness satisfaction score for the item based at least in part on the set of item data for the item and a set of user data for the user, wherein the machine-learning model derives freshness preferences of the user from the historical conversion information in the set of user data for the user;
-generating the user interface comprising the information describing the one or more items, the interface including descriptions of the one or more items and a visual indication of the user-specific freshness satisfaction score for each of the one or more items; and
-sending the user interface to the client device associated with the user, wherein sending the user interface causes the client device to display the user interface
These limitations are not indicative of integration into a practical application because:
The additional elements of claim 1 are recited at a high level of generality (i.e. as generic computing hardware) such that they amount to nothing more than mere instructions to implement or apply the abstract idea on a generic computing hardware (or, merely use a computer as a tool to perform an abstract idea) as supported by paragraph [0090] of Applicant’s specification – “Any of the steps, operations, or processes described herein may be performed or implemented with one or more hardware or software modules, alone or in combination with other devices. In some embodiments, a software module is implemented with a computer program product comprising one or more computer-readable media storing computer program code or instructions, which can be executed by a computer processor for performing any or all of the steps, operations, or processes described.” Specifically, the additional elements of a computer system, a processor, a computer-readable medium, a client device, an online concierge system, a user interface, a machine-learning model, the model being trained, and sending the user interface causes the client device to display the user interface are recited at a high-level of generality (i.e. as a generic processor performing the generic computer functions of receiving data, retrieving data, accessing data, training, applying data, generating data, and sending data) such that they amount do no more than mere instructions to apply the exception using generic computer components. Accordingly, these additional elements do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea. The claim is directed to an abstract idea. Further, the additional elements do no more than generally link the use of the judicial exception to a particular technological environment or field of use (such as computers or computing networks). Employing well-known computer functions to execute an abstract idea, even when limiting the use of the idea to one particular environment, does not integrate the exception into a practical application.

Additionally, the additional elements are insufficient to integrate the abstract idea into a practical application because the claim fails to i) reflect an improvement in the functioning of a computer or an improvement to another technology or technical field, ii) apply the judicial exception with, or use the judicial exception in conjunction with, a particular machine or manufacture that is integral to the claim, iii) effect a transformation or reduction of a particular article to a different state or thing, or iv) apply or use the judicial exception in some other meaningful way beyond generally linking the use of the judicial exception to a particular technological environment.
Accordingly, the judicial exception is not integrated into a practical application.

Under Step 2B, it is determined whether the claims recite additional elements that amount to significantly more than the judicial exception. The claims of the present application do not include additional elements that are sufficient to amount to significantly more than the judicial exception (NO).

In the case of claim 1, taken individually or as a whole, the additional elements of claim 1 do not provide an inventive concept. As discussed above under step 2A (prong 2) with respect to the integration of the abstract idea into a practical application, the additional elements used to perform the claimed functions amount to no more than a general link to a technological environment.

Even considered as an ordered combination (as a whole), the additional elements do not add anything significantly more than when considered individually.

Claim 11 is a computer program product reciting similar functions as claim 1. Examiner notes that claim 11 recites the additional elements of a computer program, a non-transitory computer-readable storage medium having instructions encoded, a processor, a client device, an online concierge system, a user interface, a machine-learning model, the model being trained, and sending the user interface causes the client device to display the user interface, however, claim 11 does not qualify as eligible subject matter for similar reasons as claim 1 indicated above.

Claim 20 is a computer system reciting similar functions as claim 1. Examiner notes that claim 20 recites the additional elements of computer system, a non-transitory computer-readable storage medium storing instructions, a processor, a client device, an online concierge system, a user interface, a machine-learning model, the model being trained, and sending the user interface causes the client device to display the user interface, however, claim 20 does not qualify as eligible subject matter for similar reasons as claim 1 indicated above.

Therefore, claims 11 and 20 do not provide an inventive concept and do not qualify as eligible subject matter.

Dependent claims 4-10 and 14-19, when analyzed as a whole, are held to be patent ineligible under 35 U.S.C. § 101 because they do not add “significantly more” to the abstract idea. More specifically, dependent claims 4-10 and 14-19 further fall within the “Certain Methods of Organizing Human Activity” grouping of abstract ideas in that they recite commercial interactions. Dependent claim 10 does not recite any farther additional elements, and as such are not indicative of integration into a practical application for at least similar reasons discussed above. Dependent claims 4-9 and 14-19 recite the additional elements of the machine-learning model being further trained, the machine-learning model, and the user interface but similar to the analysis under prong two of Step 2A these additional elements are used as a tool to perform the abstract idea. As such, under prong two of Step 2A, claims 4-10 and 14-19 are not indicative of integration into a practical application for at least similar reasons as discussed above. Thus, dependent claims 4-10 and 14-19 are “directed to” an abstract idea. Next, under Step 2B, similar to the analysis of claims 1, 11 and 20, dependent claims 4-10 and 14-19 when analyzed individually and as an ordered combination, merely further define the commonplace business method (i.e. determining freshness satisfaction scores of items for purchase and providing the item data to a user) being applied on a general-purpose computer and, therefore, do not amount to significantly more than the abstract idea itself. Accordingly, the Examiner concludes that there are no meaningful limitations in the claims that transform the judicial exception into a patent eligible application such that the claims amount to significantly more than the judicial exception itself. The analysis above applies to all statutory categories of invention.

Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.

The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1, 4, 6, 9-10, 14, 16, and 19-20, are rejected under 35 U.S.C. 103 as being unpatentable over Hsu et al (US 2018/0047071 A1), hereinafter Hsu, in view of Krishnan Gorumkonda et al. (US 9,824,298 B1), hereinafter Krishnan Gorumkonda.
Regarding claim 1, Hsu discloses a method, performed at a computer system comprising a processor and a computer-readable medium, comprising:
-receiving, from a client device associated with a user of an online concierge system, a request to access a user interface comprising information describing one or more items included among an inventory at a retailer location (Hsu, see at least: “FIGS. 10-13 are illustrative “screen shots” or displays, showing how features of an embodiment of the invention may be presented to a consumer [i.e. from a client device associated with a user of an online concierge system]. Referring to FIG. 10, in one embodiment of the inventive system and methods, a combined aggregate review (CAR) score or predictive rating 1002 for a base product 1004 is associated with the base product's variants and presented to consumers via a searchable website [i.e. receiving a request to access a user interface comprising information describing one or more items included among an inventory at a retailer location]” [0217] and “while the specific embodiments of the invention described herein are associated with goods that may be purchased in a store or via a website [i.e. one or more items included among an inventory at a retailer location]” [0072] and Fig. 10 shows the interface of the searchable webpage includes a results webpage for the search query of ‘cameras’ [i.e. receiving a request to access a user interface] and the resulting items from the search [i.e. a user interface comprising information describing one or more items included among an inventory at a retailer location]);
-retrieving a set of item data for an item included among the inventory at the retailer location (Hsu, see at least: “user reviews may be collected as part of a web page “crawling” process and may be used as part of collecting data about a merchant's catalog [i.e. retrieving a set of item data for an item included among the inventory at the retailer location]. Alternately, URLs for review pages for a merchant can be generated from known SKU's for that merchant, where such information may have been collected via an alternate means (e.g., affiliate feed ingestion)” [0100] and “while the specific embodiments of the invention described herein are associated with goods that may be purchased in a store or via a website [i.e. at the retailer location]” [0072]);
-accessing a machine-learning model trained to predict a satisfaction score for the item, wherein the satisfaction score indicates a measure of satisfaction of the user with a quality of the item (Hsu, see at least: “As shown in the figure, historical product ratings and other product data (stage 904) may be used to generate training data (stage 908) [i.e. accessing a machine-learning model trained] that can be used to generate models (stage 912). To generate current predictions for a product catalog, the system first generates predictive features for each product based on information known about the product currently (stage 916), and then applies the model learned in stage 912 to create the “prediction” (stage 920)” [0167] and “machine learning techniques and methods can be used to generate a predictor of future aggregate product ratings from historical data on product sales, ratings, and reviews. Given a database of existing products, including data related to both externally generated product scores as well as structured product or product-related information (e.g., brand reputation, base model quality, base model and variant model features, average price level, etc.), a machine learning problem can be formulated to “predict” the expected future rating or rank [i.e. wherein the satisfaction score indicates a measure of satisfaction of the user with a quality of the item] of a product based on known information” [0070] and “Rating generation (transformation of prediction into rating) 420: this component is responsible for translating the prediction generated by the machine learning model into a more easily understandable score [i.e. to predict a satisfaction score for the item], rating, ranking, etc.” [0090]) and the machine learning model is trained by:
-receiving item data for a plurality of items (Hsu, see at least: “historical product ratings and other product data (stage 904) may be used to generate training data (stage 908) [i.e. the machine-learning model is trained by: receiving item data] that can be used to generate models (stage 912)” [0167] and “user reviews may be collected as part of a web page “crawling” process and may be used as part of collecting data about a merchant's catalog [i.e. for a plurality of items]. Alternately, URLs for review pages for a merchant can be generated from known SKU's for that merchant, where such information may have been collected via an alternate means (e.g., affiliate feed ingestion)” [0100]),
-receiving user data for a plurality of users, the user data including historical conversion information associated with each user of the plurality of users (Hsu, see at least: “historical product ratings [i.e. receiving user data for a plurality of users, the user data including historical conversion information associated with each user of the plurality of users] and other product data (stage 904) may be used to generate training data (stage 908) that can be used to generate models (stage 912)” [0167] and “One or more of the techniques described herein may be applied to a “domain” where reviews and ratings are generated by expert sources and/or by consumers. One or more of the methods, functions, processes, or operations described herein relating to combining expert and consumer review ratings may be applied to a domain where consumer reviews/ratings can be collected individually or in aggregate [i.e. for a plurality of users]” [0071] Examiner notes that conversion data that ‘conversion data’ is interpreted in light of paragraph [0037] of Applicant’s spec which describes “the data collection module 200 collects conversion data, such as order data or purchase data … Order data may further include information describing … a rating that the user gave the order … or a review”),
-receiving conversion data for a plurality of conversions by the plurality of users (Hsu, see at least: “historical product ratings [i.e. receiving conversion data for a plurality of conversions] and other product data (stage 904) may be used to generate training data (stage 908) that can be used to generate models (stage 912)” [0167] and “One or more of the techniques described herein may be applied to a “domain” where reviews and ratings are generated by expert sources and/or by consumers. One or more of the methods, functions, processes, or operations described herein relating to combining expert and consumer review ratings may be applied to a domain where consumer reviews/ratings can be collected individually or in aggregate [i.e. by a plurality of users]” [0071] Examiner notes that conversion data that ‘conversion data’ is interpreted in light of paragraph [0037] of Applicant’s spec which describes “the data collection module 200 collects conversion data, such as order data or purchase data … Order data may further include information describing … a rating that the user gave the order … or a review”);
-receiving, for each conversion of the plurality of conversions, a label describing the measure of satisfaction of a corresponding user with the quality of a set of items associated with the conversion (Hsu, see at least: “A variety of available data can be used as a predictive feature in the prediction model, including but not limited to: … Extracted sentiments for the product based on reviews [i.e. receiving, for each conversion of the plurality of conversions, a label]” [0168] and [0178] and “Implicit “ratings” may be identified in a variety of ways, including using keywords, using learned attributes and values, or by using more advanced parsing and textual analysis techniques. Additionally, for such implicit dimensions, a properly constructed system can estimate a user rating based on heuristics, or more generally by training a machine learning classifier to estimate how positive or negative a user views a product with respect to the dimension in question (i.e., the user's positive/negative sentiment about the dimension) [i.e. a label describing the measure of satisfaction of a corresponding user with the quality of a set of items associated with the conversion]” [0201]), and
-adjusting parameters of the machine-learning model based at least in part on the item data, the user data for the plurality of users, the conversion data, and the label for each conversion of the plurality of conversions (Hsu, see at least: “the number of purchases of a product based on the number of reviews may be estimated by modeling the stream of written reviews using a dynamic Bayesian network (for example, a Kalman Filter or switching Kalman Filter) that models the joint distribution of written reviews over time and accounts for estimated variables such as the number of purchases, where the observable variable represents the number of reviews written in a time period and the hidden variable(s) represent the number of people that purchase the product during the same time period and the number of people who have yet to write a review. The parameters of this Bayesian network (e.g., the probability of a purchasing consumer writing a review, the change in the rate of purchases over time) can be trained [i.e. adjusting parameters of the machine-learning model] via an expectation-maximization algorithm over a dataset that comprises review count sequences for different products” [0142] and “historical product ratings and other product data (stage 904) [i.e. the item data, the user data for the plurality of users, the conversion data] may be used to generate training data (stage 908) that can be used to generate models (stage 912)” [0167] and “A variety of available data can be used as a predictive feature in the prediction model [i.e. adjusting parameters of the machine-learning model based at least in part on], including but not limited to: … Extracted sentiments for the product based on reviews [i.e. the user data for the plurality of users, the label for each conversion of the plurality of conversions]” [0168] and [0178] and “Implicit “ratings” may be identified in a variety of ways, including using keywords, using learned attributes and values, or by using more advanced parsing and textual analysis techniques. Additionally, for such implicit dimensions, a properly constructed system can estimate a user rating based on heuristics, or more generally by training a machine learning classifier [i.e. adjusting parameters of the machine-learning model based at least in part on] to estimate how positive or negative a user views a product with respect to the dimension in question (i.e., the user's positive/negative sentiment about the dimension) [i.e. the label for each conversion of the plurality of conversions]” [0201]);
-applying the machine-learning model to predict the satisfaction score for the item based at least in part on the set of item data for the item and a set of user data, wherein the machine-learning model derives preferences of the user from the historical conversion information in the set of user data (Hsu, see at least: “As shown in the figure, historical product ratings [i.e. and a set of user data] and other product data (stage 904) may be used to generate training data (stage 908) that can be used to generate models (stage 912). To generate current predictions for a product catalog, the system first generates predictive features for each product based on information known about the product currently (stage 916) [i.e. for the item based at least in part on the set of item data for the item], and then applies the model learned in stage 912 to create the “prediction” (stage 920) [i.e. applying the machine-learning model to predict the satisfaction score]” [0167] and “A variety of available data can be used as a predictive feature in the prediction model, including but not limited to: … Extracted sentiments for the product based on reviews [i.e. and a set of user data, wherein the machine-learning model derives preferences of the user from the historical conversion information in the set of user data]” [0168] and [0178]);
-generating the user interface comprising the information describing the one or more items, the interface including descriptions of the one or more items and a visual indication of the satisfaction score for each of the one or more items (Hsu, see at least: “FIGS. 10-13 are illustrative “screen shots” or displays, showing how features of an embodiment of the invention may be presented to a consumer. Referring to FIG. 10, in one embodiment of the inventive system and methods, a combined aggregate review (CAR) score or predictive rating 1002 [i.e. generating the user interface comprising the information describing the one or more items, the interface including descriptions of the one or more items and a visual indication of the satisfaction score for each of the one or more items] for a base product 1004 is associated with the base product's variants and presented to consumers via a searchable website” [0217] and Fig. 10 shows the interface [i.e. generating the user interface comprising] of the searchable webpage includes a results webpage for the search query of ‘cameras’ that comprise the generated predictive ratings associated with each search result and item descriptions of each data [i.e. the information describing the one or more items, the interface including descriptions of the one or more items and a visual indication of the satisfaction score for each of the one or more items]); and
-sending the user interface to the client device associated with the user, wherein sending the user interface causes the client device to display the user interface (Hsu, see at least: “FIGS. 10-13 are illustrative “screen shots” or displays, showing how features of an embodiment of the invention may be presented to a consumer [i.e. sending the user interface to the client device associated with the user, wherein sending the user interface causes the client device to display the user interface]. Referring to FIG. 10, in one embodiment of the inventive system and methods, a combined aggregate review (CAR) score or predictive rating 1002 for a base product 1004 is associated with the base product's variants and presented to consumers via a searchable website” [0217] and Fig. 10).

Hsu does not explicitly disclose, responsive to receiving the request, retrieving a set of item data for an item included among the inventory at the retailer location, the set of item data including freshness information describing a current state of a freshness lifecycle of the item and current environmental information associated with a location of the item at the retailer location; the satisfaction score being a user-specific freshness satisfaction score, wherein the freshness satisfaction score indicates a measure of satisfaction of the user with a freshness of the item; the item data including freshness information about a freshness lifecycle of each item and information describing environmental information associated with the locations of the items; the plurality of users being a plurality of users of the online concierge system; the measure of satisfaction of a corresponding user being with the freshness of each item in a set of items associated with the conversion; item data including the freshness information about a freshness lifecycle of each item and the information about environmental information associated with the locations of the items; predicting the user-specific freshness satisfaction score for the item based at least in part on the set of item data for the item and a set of user data for the user; preferences being freshness preferences and the set of user data being for the user; the satisfaction score being a user-specific freshness satisfaction score.
Krishnan Gorumkonda, however, teaches predicting and detecting product quality (i.e. abstract), including the known technique of, responsive to receiving the request, retrieving a set of item data for an item included among the inventory at the retailer location, the set of item data including freshness information describing a current state of a freshness lifecycle of the item and current environmental information associated with a location of the item at the retailer location (Krishnan Gorumkonda, see at least: “The process 100 may also determine an item that corresponds with the request at 130 [i.e. responsive to receiving the request]. For example, the computing device 104 may receive a selection or identification of a desired ripeness score in a communication from the user (e.g., via the network page 124, via messaging system external to the network page, etc.). As illustrated, the user may choose the first image 126 associated with the first ripeness score. The computing device 104 may determine a corresponding produce item at a facility (e.g., fulfillment facility, merchant facility, third party facility, virtual management facility, etc.) [i.e. retrieving a set of item data for an item included among the inventory at the retailer location] associated with the first ripeness score (e.g., based in part on the training data set, an extrapolated ripeness score of one or more produce items in the facility, and/or an identification of a produce item that should achieve the ripeness score by the specified date) [i.e. the set of item data including freshness information describing a current state of a freshness lifecycle of the item]” Col. 4 Ln. 13-27 and “The system checks the apples for firmness (e.g., wrinkles, texture, etc.) utilizing one or more sensors (e.g., imaging devices, cameras, etc.), and data generated by the sensors (e.g., captured images from imaging device(s) or camera(s) including infrared (IR) images, digital (red/green/blue or RGB) images, subsurface scattering, etc.) to identify other characteristics of the apples. The system compares the data utilized to identify visual and/or infrared characteristics with other apples that were analyzed earlier in order to determine the relative ripeness of each apple of the crate of apples [i.e. and current environmental information associated with a location of the item at the retailer location]. Apples from the crate and other crates are then offered to users to order through a network page. When the user requests apples at the ripeness level that matches the ripeness level of at least some of the apples from the crate of apples, apples from the crate can be shipped to the user” Col. 2 Ln. 21-35);
the known technique of predicting a user-specific freshness satisfaction score for the item, wherein the freshness satisfaction score indicates a measure of satisfaction of the user with a freshness of the item (Krishnan Gorumkonda, see at least: “a system for predicting and detecting produce quality (e.g., a ripeness score of a produce item [i.e. predict a user-specific freshness satisfaction score for the item], whether mold is present, etc.). For example, data representing visual or infrared characteristics of a produce item (e.g., skin, shape, wrinkles, and other characteristics of an apple, pepper, etc.) may be compared with ripeness characteristics of the type of produce item (e.g., other apples or peppers). The ripeness characteristics may correspond with the type of produce item at different stages of ripeness along a ripeness regression (e.g., a timeline of the produce item from raw to rotten). One or more ripeness scores of the produce item may be determined along a timeline (e.g., raw at day 1, ripe at day 5, rotten at day 10, etc.), so that when a user requests a produce item corresponding with a particular ripeness score (e.g., to be ripe on a particular date), the produce item can be provided to the user based in part on the data representing visual or infrared characteristics of the produce item and ripeness regression [i.e. wherein the freshness satisfaction score indicates a measure of satisfaction of the user with a freshness of the item]” Col. 2 Ln. 1-18 and “the user may specify Saturday morning on the calendar 612, corresponding with the time that the user plans to bake banana bread (e.g., using the very ripe bananas) [i.e. a user-specific freshness satisfaction score]. The user may access the calendar 612 to identify Saturday morning as the desired date to receive the produce item” Col. 17 Ln. 5-10; Examiner notes that the ripeness score is a user-specific freshness satisfaction score as the desired ripeness is based on a preference of a user (e.g. one user will be satisfied with ripeness score of 8 for banana bread and a different user would be satisfied with a score of 4 for snacking) [i.e. predict a user-specific freshness satisfaction score]);
the known technique of the item data including freshness information about a freshness lifecycle of each item and information describing environmental information associated with the locations of the items (Krishnan Gorumkonda, see at least: “the item data may include … origin or source data 114 (e.g., the source that provided the produce item, including a state, city, farm, climate associated with a location, soil composition used to grow the produce item, or other environmental factors that may affect the ripeness regression of the produce item) [i.e. the item data including information describing environmental information associated with the locations of the items]” Col. 2 Ln. 53-67 & Col. 3 Ln. 1 and “The ripeness characteristics of those particular oranges may be imparted on the new produce item from the similar origin. In some examples, the ripeness characteristics can include a visual representation of ripeness regression associated with the particular produce type [i.e. the item data including freshness information about a freshness lifecycle of each item]. The ripeness regression may correspond with at least one of a time frame (e.g., day 1 to day 5) or one or more environmental factors (e.g., heat, humidity, light, etc.) [i.e. and information describing environmental information associated with the locations of the items]” Col. 10 Ln. 10-18 and “the environmental factors may be associated with the transportation and storage of the produce item, including transportation by various vehicles, in various types of packaging, over a particular time frame and/or storing the item in a refrigerator, in a freezer, on a shelf, in the sun, etc. [i.e. and information describing environmental information associated with the locations of the items]” Col. 10 Ln. 49-54);
the known technique of receiving user data for a plurality of users of the online concierge system (Krishnan Gorumkonda, see at least: “The feedback module 246 may be configured to receive feedback. For example, the user may confirm the ripeness of the produce item when it is received with their desired ripeness score, or provide additional item data regarding the ripeness of the item at arrival [i.e. for a plurality of users of the online concierge system]. The feedback module 246 may also be configured to update item data (e.g., via the item data module 236 or training data module 238) to account for the additional information from the user and/or update the ripeness regression (e.g., via the regression module 240) to identify the ripeness of the produce item at a particular date” Col. 8 Ln. 55-64 and “feedback 112 (e.g., communication(s) from the user identifying the received ripeness when the produce item arrived at a location associated with the user, including “This was too ripe when I needed it!” or “Perfect! Thanks!”) [i.e. receiving user data for a plurality of users]” Col. 2 Ln. 60-63);
the known technique of receiving conversion data for a plurality of conversions by a plurality of users of the online concierge system (Krishnan Gorumkonda, see at least: “The feedback module 246 may be configured to receive feedback. For example, the user may confirm the ripeness of the produce item when it is received with their desired ripeness score, or provide additional item data regarding the ripeness of the item at arrival [i.e. by a plurality of users of the online concierge system]. The feedback module 246 may also be configured to update item data (e.g., via the item data module 236 or training data module 238) to account for the additional information from the user and/or update the ripeness regression (e.g., via the regression module 240) to identify the ripeness of the produce item at a particular date” Col. 8 Ln. 55-64 and “feedback 112 (e.g., communication(s) from the user identifying the received ripeness when the produce item arrived at a location associated with the user, including “This was too ripe when I needed it!” or “Perfect! Thanks!”) [i.e. receiving conversion data for a plurality of conversions]” Col. 2 Ln. 60-63 Examiner notes that conversion data that ‘conversion data’ is interpreted in light of paragraph [0037] of Applicant’s spec which describes “the data collection module 200 collects conversion data, such as order data or purchase data … Order data may further include information describing … a rating that the user gave the order … or a review”);
the known technique of a measure of satisfaction of a corresponding user with the freshness of each item in a set of items associated with the conversion (Krishnan Gorumkonda, see at least: “a system for predicting and detecting produce quality (e.g., a ripeness score of a produce item, whether mold is present, etc.). For example, data representing visual or infrared characteristics of a produce item (e.g., skin, shape, wrinkles, and other characteristics of an apple, pepper, etc.) may be compared with ripeness characteristics of the type of produce item (e.g., other apples or peppers). The ripeness characteristics may correspond with the type of produce item at different stages of ripeness along a ripeness regression (e.g., a timeline of the produce item from raw to rotten). One or more ripeness scores of the produce item may be determined along a timeline (e.g., raw at day 1, ripe at day 5, rotten at day 10, etc.), so that when a user requests a produce item corresponding with a particular ripeness score (e.g., to be ripe on a particular date), the produce item can be provided to the user based in part on the data representing visual or infrared characteristics of the produce item and ripeness regression [i.e. a measure of satisfaction of a corresponding user with the freshness of each item in a set of items]” Col. 2 Ln. 1-18 and “The process 100 can begin with accessing item data at 102. For example, a computing device 104 can interact with a data store 106 to access stored data representing visual or infrared characteristics of an item, item data, and/or interact with one or more items [i.e. of each item in a set of items] to determine the item data. For example, the item data may include image data 108 (e.g., image data/photographs of the produce item [i.e. each item in a set of items], other images from one or more angles, etc.), ripeness data 110 (e.g., “raw” corresponds with a ripeness score of 0-2, “ripe” corresponds with a ripeness score of 3-7, “rotten” corresponds with a ripeness score of 8-10, etc., such that a green banana is 2, a yellow banana is 5 and a brown banana is 8)” Col. 2 Ln. 48-60 and “feedback 112 (e.g., communication(s) from the user identifying the received ripeness when the produce item arrived at a location associated with the user, including “This was too ripe when I needed it!” or “Perfect! Thanks!”) [i.e. associated with the conversion]” Col. 2 Ln. 60-63);
the known technique of item data including the freshness information about a freshness lifecycle of each item and the information about environmental information associated with the locations of the items (Krishnan Gorumkonda, see at least: “the item data may include … origin or source data 114 (e.g., the source that provided the produce item, including a state, city, farm, climate associated with a location, soil composition used to grow the produce item, or other environmental factors that may affect the ripeness regression of the produce item) [i.e. item data including the information about environmental information associated with the locations of the items]” Col. 2 Ln. 53-67 & Col. 3 Ln. 1 and “The ripeness characteristics of those particular oranges may be imparted on the new produce item from the similar origin. In some examples, the ripeness characteristics can include a visual representation of ripeness regression associated with the particular produce type [i.e. the item data including the freshness information about a freshness lifecycle of each item]. The ripeness regression may correspond with at least one of a time frame (e.g., day 1 to day 5) or one or more environmental factors (e.g., heat, humidity, light, etc.) [i.e. and the information describing environmental information associated with the locations of the items]” Col. 10 Ln. 10-18 and “the environmental factors may be associated with the transportation and storage of the produce item, including transportation by various vehicles, in various types of packaging, over a particular time frame and/or storing the item in a refrigerator, in a freezer, on a shelf, in the sun, etc. [i.e. and information describing environmental information associated with the locations of the items]” Col. 10 Ln. 49-54);
the known technique of predicting the user-specific freshness satisfaction score for the item based at least in part on the set of item data for the item and a set of user data for the user (Krishnan Gorumkonda, see at least: “the user may specify Saturday morning on the calendar 612, corresponding with the time that the user plans to bake banana bread (e.g., using the very ripe bananas) [i.e. the user-specific freshness satisfaction score for the item]. The user may access the calendar 612 to identify Saturday morning as the desired date to receive the produce item” Col. 17 Ln. 5-10 and “The request module 242 may be configured to receive a communication from a user that identifies a desired ripeness level [i.e. the user-specific freshness satisfaction score for the item] (e.g., based on data representing visual or infrared characteristics of the produce item, based on a time frame to use the produce item, based on the purpose of the produce item, etc.) [i.e. predict the user-specific freshness satisfaction score for the item based at least in part on the set of item data for the item]. In some examples, the request module 242 may identify the request without a communication from a user (e.g., based on an identified pattern in a user's order history, based on browsing characteristics of the user [i.e. and a set of user data for the user], etc.)” Col. 8 Ln. 30-39 and “the ripeness score of the produce item is determined from a passive communication (e.g., shopping/order history, browsing history, interactions with network pages, etc.) [i.e. and a set of user data for the user]” Col. 10 Ln. 38-41),
the known technique of deriving freshness preferences of the user from the historical conversion information in the set of user data for the user (Krishnan Gorumkonda, see at least: “The request module 242 may be configured to receive a communication from a user that identifies a desired ripeness level (e.g., based on data representing visual or infrared characteristics of the produce item, based on a time frame to use the produce item, based on the purpose of the produce item, etc.). In some examples, the request module 242 may identify the request without a communication from a user (e.g., based on an identified pattern in a user's order history, based on browsing characteristics of the user [i.e. deriving freshness preferences of the user from the historical conversion information in the set of user data for the user], etc.)” Col. 8 Ln. 30-39 and “the ripeness score of the produce item is determined from a passive communication (e.g., shopping/order history, browsing history, interactions with network pages, etc.) [i.e. from the historical conversion information in the set of user data for the user]” Col. 10 Ln. 38-41 Examiner notes that conversion data that ‘conversion data’ is interpreted in light of paragraph [0037] of Applicant’s spec which describes “the data collection module 200 collects conversion data, such as order data or purchase data … Order data may further include information describing … a rating that the user gave the order … or a review”); and
the known technique of the user-specific freshness satisfaction score (Krishnan Gorumkonda, see at least: “a system for predicting and detecting produce quality (e.g., a ripeness score of a produce item, whether mold is present, etc.). For example, data representing visual or infrared characteristics of a produce item (e.g., skin, shape, wrinkles, and other characteristics of an apple, pepper, etc.) may be compared with ripeness characteristics of the type of produce item (e.g., other apples or peppers). The ripeness characteristics may correspond with the type of produce item at different stages of ripeness along a ripeness regression (e.g., a timeline of the produce item from raw to rotten). One or more ripeness scores of the produce item may be determined along a timeline (e.g., raw at day 1, ripe at day 5, rotten at day 10, etc.), so that when a user requests a produce item corresponding with a particular ripeness score (e.g., to be ripe on a particular date) [i.e. a user-specific freshness satisfaction score], the produce item can be provided to the user based in part on the data representing visual or infrared characteristics of the produce item and ripeness regression” Col. 2 Ln. 1-18 and “the user may specify Saturday morning on the calendar 612, corresponding with the time that the user plans to bake banana bread (e.g., using the very ripe bananas) [i.e. a user-specific freshness satisfaction score]. The user may access the calendar 612 to identify Saturday morning as the desired date to receive the produce item” Col. 17 Ln. 5-10; Examiner notes that the ripeness score is a user-specific freshness satisfaction score as the desired ripeness is based on a preference of a user (e.g. one user will be satisfied with ripeness score of 8 for banana bread and a different user would be satisfied with a score of 4 for snacking) [i.e. a user-specific freshness satisfaction score]). These known techniques are applicable to the method of Hsu as they both share characteristics and capabilities, namely, they are directed to predicting and detecting product quality.
It would have been recognized that applying the known techniques of responsive to receiving the request, retrieving a set of item data for an item included among the inventory at the retailer location, the set of item data including freshness information describing a current state of a freshness lifecycle of the item and current environmental information associated with a location of the item at the retailer location; predicting a user-specific freshness satisfaction score for the item, wherein the freshness satisfaction score indicates a measure of satisfaction of the user with a freshness of the item; the item data including freshness information about a freshness lifecycle of each item and information describing environmental information associated with the locations of the items; receiving user data for a plurality of users of the online concierge system; receiving conversion data for a plurality of conversions by a plurality of users of the online concierge system; a measure of satisfaction of a corresponding user with the freshness of each item in a set of items associated with the conversion; item data including the freshness information about a freshness lifecycle of each item and the information about environmental information associated with the locations of the items; predicting the user-specific freshness satisfaction score for the item based at least in part on the set of item data for the item and a set of user data for the user, deriving freshness preferences of the user from the historical conversion information in the set of user data for the user; and the user-specific freshness satisfaction score, as taught by Krishnan Gorumkonda, to the teachings of Hsu would have yielded predictable results because the level of ordinary skill in the art demonstrated by the references applied shows the ability to incorporate such references into similar methods. Further, adding the modifications of responsive to receiving the request, retrieving a set of item data for an item included among the inventory at the retailer location, the set of item data including freshness information describing a current state of a freshness lifecycle of the item and current environmental information associated with a location of the item at the retailer location; predicting a user-specific freshness satisfaction score for the item, wherein the freshness satisfaction score indicates a measure of satisfaction of the user with a freshness of the item; the item data including freshness information about a freshness lifecycle of each item and information describing environmental information associated with the locations of the items; receiving user data for a plurality of users of the online concierge system; receiving conversion data for a plurality of conversions by a plurality of users of the online concierge system; a measure of satisfaction of a corresponding user with the freshness of each item in a set of items associated with the conversion; item data including the freshness information about a freshness lifecycle of each item and the information about environmental information associated with the locations of the items; predicting the user-specific freshness satisfaction score for the item based at least in part on the set of item data for the item and a set of user data for the user, deriving freshness preferences of the user from the historical conversion information in the set of user data for the user; and the user-specific freshness satisfaction score, as taught by Krishnan Gorumkonda, into the method of Hsu would have been recognized by those of ordinary skill in the art as resulting in an improved method that would provide a user with a desired ripeness of produce (Krishnan Gorumkonda, Col. 2 Ln. 1-18).

Regarding claim 4, Hsu in view of Krishnan Gorumkonda teaches the method of claim 1. Hsu further discloses:
-wherein applying the machine-learning model to predict the satisfaction score for the item based at least in part on the set of item data for the item and a set of user data comprises applying the machine-learning model to predict the satisfaction score for the item based at least in part on one or more of: information describing a seasonality of the item, information describing an availability of the item, environmental information associated with the item, historical conversion information associated with the item, or information describing a life cycle of the item (Hsu, see at least: “As shown in the figure, historical product ratings [i.e. and a set of user data] and other product data (stage 904) may be used to generate training data (stage 908) that can be used to generate models (stage 912). To generate current predictions for a product catalog, the system first generates predictive features for each product based on information known about the product currently (stage 916) [i.e. for the item based at least in part on the set of item data for the item], and then applies the model learned in stage 912 to create the “prediction” (stage 920) [i.e. applying the machine-learning model to predict the satisfaction score for the item]” [0167] and “Predictive feature generation 408: this component is responsible for generating features from a historical stream of product review [i.e. comprises applying the machine-learning model to predict the satisfaction score for the item based at least in part on one or more of: information describing a seasonality of the item, information describing an availability of the item, environmental information associated with the item, historical conversion information associated with the item, or information describing a life cycle of the item] and product data that the machine learning system will then use to “predict” the desired target (as defined in 404)” [0087]).
Hsu does not explicitly disclose the satisfaction score being a freshness satisfaction score and a set of user data being and a set of user data for the user.
Krishnan Gorumkonda, however, teaches predicting and detecting product quality (i.e. abstract), including the known technique of a freshness satisfaction score and a set of user data for the user (Krishnan Gorumkonda, see at least: “a system for predicting and detecting produce quality (e.g., a ripeness score of a produce item, whether mold is present, etc.). For example, data representing visual or infrared characteristics of a produce item (e.g., skin, shape, wrinkles, and other characteristics of an apple, pepper, etc.) may be compared with ripeness characteristics of the type of produce item (e.g., other apples or peppers). The ripeness characteristics may correspond with the type of produce item at different stages of ripeness along a ripeness regression (e.g., a timeline of the produce item from raw to rotten). One or more ripeness scores of the produce item may be determined along a timeline (e.g., raw at day 1, ripe at day 5, rotten at day 10, etc.), so that when a user requests a produce item corresponding with a particular ripeness score (e.g., to be ripe on a particular date) [i.e. a freshness satisfaction score], the produce item can be provided to the user based in part on the data representing visual or infrared characteristics of the produce item and ripeness regression” Col. 2 Ln. 1-18 and “the ripeness score of the produce item is determined from a passive communication (e.g., shopping/order history, browsing history, interactions with network pages, etc.) [i.e. a set of user data for the user]” Col. 10 Ln. 38-41). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Hsu with Krishnan Gorumkonda for the reasons identified above with respect to claim 1.

Regarding claim 6, Hsu in view of Krishnan Gorumkonda teaches the method of claim 4. Hsu further discloses:
-wherein applying the machine-learning model to predict the satisfaction score for the item based at least in part on historical conversion information associated with the item comprises applying the machine-learning model to predict the satisfaction score for the item based at least in part on one or more of: a time associated with a previous conversion associated with the item or user data associated with a user associated with a previous conversion associated with the item (Hsu, see at least: “As shown in the figure, historical product ratings and other product data (stage 904) may be used to generate training data (stage 908) that can be used to generate models (stage 912). To generate current predictions for a product catalog, the system first generates predictive features for each product based on information known about the product currently (stage 916) [i.e. for the item based at least in part on the set of item data for the item], and then applies the model learned in stage 912 to create the “prediction” (stage 920) [i.e. applying the machine-learning model to predict the satisfaction score for the item]” [0167] and “A machine learning system for predicting future review aggregates operates in a way that is similar to the candidate ranking/rating method evaluation described previously. In one embodiment, the system takes as inputs: (1) An “ideal” target function with the description and restrictions described previously; and (2) time-stamped product review data [i.e. applying the machine-learning model to predict the satisfaction score for the item based at least in part on one or more of: a time associated with a previous conversion associated with the item or user data associated with a user associated with a previous conversion associated with the item] (or other relevant data)” [0166]).
Hsu does not explicitly disclose the satisfaction score being a freshness satisfaction score.
Krishnan Gorumkonda, however, teaches predicting and detecting product quality (i.e. abstract), including the known technique of a freshness satisfaction score (Krishnan Gorumkonda, see at least: “a system for predicting and detecting produce quality (e.g., a ripeness score of a produce item, whether mold is present, etc.). For example, data representing visual or infrared characteristics of a produce item (e.g., skin, shape, wrinkles, and other characteristics of an apple, pepper, etc.) may be compared with ripeness characteristics of the type of produce item (e.g., other apples or peppers). The ripeness characteristics may correspond with the type of produce item at different stages of ripeness along a ripeness regression (e.g., a timeline of the produce item from raw to rotten). One or more ripeness scores of the produce item may be determined along a timeline (e.g., raw at day 1, ripe at day 5, rotten at day 10, etc.), so that when a user requests a produce item corresponding with a particular ripeness score (e.g., to be ripe on a particular date) [i.e. a freshness satisfaction score], the produce item can be provided to the user based in part on the data representing visual or infrared characteristics of the produce item and ripeness regression” Col. 2 Ln. 1-18). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Hsu with Krishnan Gorumkonda for the reasons identified above with respect to claim 1.

Regarding claim 9, Hsu in view of Krishnan Gorumkonda teaches the method of claim 1. Hsu further discloses:
-wherein generating the user interface comprising the information describing the one or more items based at least in part on the satisfaction score for the item comprises:
-generating an indicator based at least in part on the satisfaction score for the item (Hsu, see at least: “FIGS. 10-13 are illustrative “screen shots” or displays, showing how features of an embodiment of the invention may be presented to a consumer. Referring to FIG. 10, in one embodiment of the inventive system and methods, a combined aggregate review (CAR) score or predictive rating 1002 [i.e. wherein generating the user interface comprising the information describing the one or more items based at least in part on the satisfaction score for the item comprises:] for a base product 1004 is associated with the base product's variants and presented to consumers via a searchable website” [0217] and Fig. 10 shows the interface of the searchable webpage includes a results webpage for the search query of ‘cameras’ and the generated predictive ratings associated with each search result [i.e. generating an indicator based at least in part on the satisfaction score for the item]); and
-generating the user interface comprising the information describing the one or more items, wherein the information describing the one or more items comprises the indicator (Hsu, see at least: “FIGS. 10-13 are illustrative “screen shots” or displays, showing how features of an embodiment of the invention may be presented to a consumer. Referring to FIG. 10, in one embodiment of the inventive system and methods, a combined aggregate review (CAR) score or predictive rating 1002 [i.e. generating the user interface comprising the information describing the one or more items, wherein the information describing the one or more items comprises the indicator] for a base product 1004 is associated with the base product's variants and presented to consumers via a searchable website” [0217] and Fig. 10 shows the interface of the searchable webpage includes a results webpage for the search query of ‘cameras’ and the generated predictive ratings associated with each search result [i.e. wherein the information describing the one or more items comprises the indicator]).
Hsu does not explicitly disclose the satisfaction score being a freshness satisfaction score.
Krishnan Gorumkonda, however, teaches predicting and detecting product quality (i.e. abstract), including the known technique of a freshness satisfaction score (Krishnan Gorumkonda, see at least: “a system for predicting and detecting produce quality (e.g., a ripeness score of a produce item, whether mold is present, etc.). For example, data representing visual or infrared characteristics of a produce item (e.g., skin, shape, wrinkles, and other characteristics of an apple, pepper, etc.) may be compared with ripeness characteristics of the type of produce item (e.g., other apples or peppers). The ripeness characteristics may correspond with the type of produce item at different stages of ripeness along a ripeness regression (e.g., a timeline of the produce item from raw to rotten). One or more ripeness scores of the produce item may be determined along a timeline (e.g., raw at day 1, ripe at day 5, rotten at day 10, etc.), so that when a user requests a produce item corresponding with a particular ripeness score (e.g., to be ripe on a particular date) [i.e. a freshness satisfaction score], the produce item can be provided to the user based in part on the data representing visual or infrared characteristics of the produce item and ripeness regression” Col. 2 Ln. 1-18). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Hsu with Krishnan Gorumkonda for the reasons identified above with respect to claim 1.

Regarding claim 10, Hsu in view of Krishnan Gorumkonda teaches the method of claim 1. Hsu further discloses:
-ranking the one or more items based at least in part on the satisfaction score for the item (Hsu, see at least: “In addition to the generated candidate rating, a ranking that is applicable to these specific points in time can be determined from the rating [i.e. ranking the one or more items based at least in part on the satisfaction score for the item]. Here a ranking is an ordering of products from best-to-worst with respect to some metric … The choice as to whether to focus on making an accurate ranking prediction or a rating prediction may ultimately depend on product considerations (i.e., how the prediction is used with respect to a product). For example, if one cares about showing the top 10 recommended products in different verticals, then evaluating the error of the predicted rank may make more sense” [0157]); and
-selecting the one or more items based at least in part on the ranking (Hsu, see at least: “The choice as to whether to focus on making an accurate ranking prediction or a rating prediction may ultimately depend on product considerations (i.e., how the prediction is used with respect to a product). For example, if one cares about showing the top 10 recommended products [i.e. selecting the one or more items based at least in part on the ranking] in different verticals, then evaluating the error of the predicted rank may make more sense” [0157]).
Hsu does not explicitly disclose the satisfaction score being a freshness satisfaction score.
Krishnan Gorumkonda, however, teaches predicting and detecting product quality (i.e. abstract), including the known technique of a freshness satisfaction score (Krishnan Gorumkonda, see at least: “a system for predicting and detecting produce quality (e.g., a ripeness score of a produce item, whether mold is present, etc.). For example, data representing visual or infrared characteristics of a produce item (e.g., skin, shape, wrinkles, and other characteristics of an apple, pepper, etc.) may be compared with ripeness characteristics of the type of produce item (e.g., other apples or peppers). The ripeness characteristics may correspond with the type of produce item at different stages of ripeness along a ripeness regression (e.g., a timeline of the produce item from raw to rotten). One or more ripeness scores of the produce item may be determined along a timeline (e.g., raw at day 1, ripe at day 5, rotten at day 10, etc.), so that when a user requests a produce item corresponding with a particular ripeness score (e.g., to be ripe on a particular date) [i.e. a freshness satisfaction score], the produce item can be provided to the user based in part on the data representing visual or infrared characteristics of the produce item and ripeness regression” Col. 2 Ln. 1-18). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Hsu with Krishnan Gorumkonda for the reasons identified above with respect to claim 1.

Claims 11, 14, 16 and 19 recite limitations directed towards a computer program product comprising a non-transitory computer-readable storage medium having instructions encoded thereon that, when executed by a processor (Hsu, see at least: “In accordance with one or more embodiments of the invention, the system, apparatus, methods, processes, functions, and/or operations described herein for the aggregation or prediction of product or service ratings/rankings may be wholly or partially implemented in the form of a set of instructions executed by one or more programmed computer processors, such as a central processing unit (CPU), controller, processor, or microprocessor. Such computer processors may be incorporated in an apparatus, server, client or other computing device operated by, or in communication with, other components of the system … The system memory 1522 and/or the fixed disk 1508 may embody a tangible computer-readable medium” [0234]), cause the processor to perform steps. The limitations recited in claims 11, 14, 16 and 19 are parallel in nature to those addressed above for claims 1-4, 6, and 9, respectively, and are therefore rejected for those same reasons set forth above in claims 1-4, 6, and 9, respectively.

Claim 20 recites limitations directed towards a computer system comprising: a processor; and a non-transitory computer-readable storage medium storing instructions that, when executed by the processor, perform actions (Hsu, see at least: “In accordance with one or more embodiments of the invention, the system, apparatus, methods, processes, functions, and/or operations described herein for the aggregation or prediction of product or service ratings/rankings may be wholly or partially implemented in the form of a set of instructions executed by one or more programmed computer processors, such as a central processing unit (CPU), controller, processor, or microprocessor. Such computer processors may be incorporated in an apparatus, server, client or other computing device operated by, or in communication with, other components of the system … The system memory 1522 and/or the fixed disk 1508 may embody a tangible computer-readable medium” [0234]), cause the system to perform operations. The limitations recited in claim 20 are parallel in nature to those addressed above for claim 1, and are therefore rejected for those same reasons set forth above in claim 1.

Claims 5, 7, 15, and 17 are rejected under 35 U.S.C. 103 as being unpatentable over Hsu, in view of Krishnan Gorumkonda, in further view of Weaver et al. (US 2024/0054503 A1), hereinafter Weaver.
Regarding claim 5, Hsu in view of Krishnan Gorumkonda teaches the method of claim 4. Hsu further discloses:
Hsu does not explicitly disclose applying the machine-learning model to predict the freshness satisfaction score for the item based at least in part on environmental information associated with the item comprises applying the machine-learning model to predict the freshness satisfaction score for the item based at least in part on one or more of: a temperature of a location associated with the item, a humidity of the location associated with the item, or a light exposure of the location associated with the item.
Krishnan Gorumkonda, however, teaches predicting and detecting product quality (i.e. abstract), including the known technique of predicting the freshness satisfaction score for the item based at least in part on environmental information associated with the item comprises predicting the freshness satisfaction score for the item based at least in part on one or more of: a temperature of a location associated with the item, a humidity of the location associated with the item, or a light exposure of the location associated with the item (Krishnan Gorumkonda, see at least: “the type of produce item is measured and/or analyzed based in part on one or more environmental factors [i.e. predicting the freshness satisfaction score for the item based at least in part on environmental information associated with the item]. For example, the training data can include item data associated with a low humidity environment and item data associated with a high humidity environment. Temperature, moisture, light, and other information that might affect the item data may also be included with the item data (e.g., one or more environmental factors, the temperature of a facility that receives the produce item, humidity of a location of a user that receives the produce item from a facility, etc.) [i.e. comprises predicting the freshness satisfaction score for the item based at least in part on one or more of: a temperature of a location associated with the item, a humidity of the location associated with the item, or a light exposure of the location associated with the item]. As an illustration, the training data associated with an apple in low humidity, 80-degree heat, and stored in a dark room might indicate particular characteristics of the apple by day ten. The training data associated with an apple in high humidity, 80-degree heat, and stored in a bright room could indicate different characteristics of the apple by day ten, and each set of training data may be stored as separate rows in the data store 502” Col. 12 Ln. 27-44). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Hsu with Krishnan Gorumkonda for the reasons identified above with respect to claim 1.
Hsu in view of Krishnan Gorumkonda does not explicitly teach applying the machine-learning model to predict the freshness satisfaction score for the item based at least in part on environmental information associated with the item comprises applying the machine-learning model to predict the freshness satisfaction score for the item based at least in part on one or more of: a temperature of a location associated with the item, a humidity of the location associated with the item, or a light exposure of the location associated with the item.
Weaver, however, teaches scoring products (i.e. abstract), including the known technique of applying the machine-learning model to predict the freshness satisfaction score for the item based at least in part on environmental information associated with the item comprises applying the machine-learning model to predict the freshness satisfaction score for the item based at least in part on one or more of: a temperature of a location associated with the item, a humidity of the location associated with the item, or a light exposure of the location associated with the item (Weaver, see at least: “A FreshScore may be generated for a given perishable by using a trained machine learning model(s) [i.e. wherein applying the machine-learning model to predict the freshness satisfaction score for the item] to process sensor data received from a sensor that is on or near the perishable during transit of the perishable [i.e. based at least in part on environmental information associated with the item comprises], in storage, as well as historical data captured for similar products over time. For example, a computing system may receive, during transit of a perishable, sensor data from a sensor that is within a threshold distance of the perishable (e.g., a sensor affixed to a pallet of apples). The received sensor data—which may represent multiple different data points of one or more parameters (e.g., temperature, humidity, etc.) measured by the sensor at multiple different times during the transit of the perishable—can be provided as input to the trained machine learning model(s) [i.e. applying the machine-learning model to predict the freshness satisfaction score for the item based at least in part on one or more of: a temperature of a location associated with the item, a humidity of the location associated with the item, or a light exposure of the location associated with the item], which may output a FreshScore relating to the freshness of the perishable” [0017]). This known technique is applicable to the method of Hsu in view of Krishnan Gorumkonda as they both share characteristics and capabilities, namely, they are directed to scoring products.
It would have been recognized that applying the known techniques of applying the machine-learning model to predict the freshness satisfaction score for the item based at least in part on environmental information associated with the item comprises applying the machine-learning model to predict the freshness satisfaction score for the item based at least in part on one or more of: a temperature of a location associated with the item, a humidity of the location associated with the item, or a light exposure of the location associated with the item, as taught by Weaver, to the teachings of Hsu in view of Krishnan Gorumkonda would have yielded predictable results because the level of ordinary skill in the art demonstrated by the references applied shows the ability to incorporate such references into similar methods. Further, adding the modification of applying the machine-learning model to predict the freshness satisfaction score for the item based at least in part on environmental information associated with the item comprises applying the machine-learning model to predict the freshness satisfaction score for the item based at least in part on one or more of: a temperature of a location associated with the item, a humidity of the location associated with the item, or a light exposure of the location associated with the item, as taught by Weaver, into the method of Hsu in view of Krishnan Gorumkonda would have been recognized by those of ordinary skill in the art as resulting in an improved method that would provide product transparency to buyers (Weaver, abstract).

Regarding claim 7, Hsu in view of Krishnan Gorumkonda teaches the method of claim 4. Hsu further discloses:
Hsu in view of Krishnan Gorumkonda does not explicitly teach applying the machine-learning model to predict the freshness satisfaction score for the item based at least in part on information describing a life cycle of the item comprises applying the machine-learning model to predict the freshness satisfaction score for the item based at least in part on one or more of: a harvest date associated with the item, a shipping and handling time associated with the item, an amount of time elapsed since the item became available for acquisition, or a shelf life associated with the item.
Weaver, however, teaches scoring products (i.e. abstract), including the known technique of applying the machine-learning model to predict the freshness satisfaction score for the item based at least in part on information describing a life cycle of the item comprises applying the machine-learning model to predict the freshness satisfaction score for the item based at least in part on one or more of: a harvest date associated with the item, a shipping and handling time associated with the item, an amount of time elapsed since the item became available for acquisition, or a shelf life associated with the item (Weaver, see at least: “the data received at block 702A may include production data indicating a batch identifier, a producer identifier, a sell-by date associated with the perishable 106 (and/or the batch), a harvest (or harvest-by) date associated with the perishable 106 (and/or the batch), a typical (e.g., average) product shelf life, a typical (e.g., average) shipping duration, average historical FreshScore for this product type and this producer, and/or a ProofScore associated with this producer and/or product, if available, and/or any other suitable type of data [i.e. based at least in part on information describing a life cycle of the item comprises applying the machine-learning model to predict the freshness satisfaction score for the item based at least in part on one or more of: a harvest date associated with the item, a shipping and handling time associated with the item, an amount of time elapsed since the item became available for acquisition, or a shelf life associated with the item]” [0073] and “At 704, the computing system 102 may provide the data received at block 702 as input to a trained machine learning model(s) 128(2) [i.e. wherein applying the machine-learning model to predict the freshness satisfaction score for the item based at least in part on information describing a life cycle of the item comprises applying the machine-learning model to]” [0076] and “At 706, the computing system 102 may generate a score(s) 104(2) associated with the perishable 106 as output from the trained machine learning model(s) 128(2). The score(s) 104(2) generated at block 706 relates to a freshness of the perishable 106 [i.e. to predict the freshness satisfaction score]” [0077]). This known technique is applicable to the method of Hsu in view of Krishnan Gorumkonda as they both share characteristics and capabilities, namely, they are directed to scoring products.
It would have been recognized that applying the known techniques of applying the machine-learning model to predict the freshness satisfaction score for the item based at least in part on information describing a life cycle of the item comprises applying the machine-learning model to predict the freshness satisfaction score for the item based at least in part on one or more of: a harvest date associated with the item, a shipping and handling time associated with the item, an amount of time elapsed since the item became available for acquisition, or a shelf life associated with the item, as taught by Weaver, to the teachings of Hsu in view of Krishnan Gorumkonda would have yielded predictable results because the level of ordinary skill in the art demonstrated by the references applied shows the ability to incorporate such references into similar methods. Further, adding the modification of applying the machine-learning model to predict the freshness satisfaction score for the item based at least in part on information describing a life cycle of the item comprises applying the machine-learning model to predict the freshness satisfaction score for the item based at least in part on one or more of: a harvest date associated with the item, a shipping and handling time associated with the item, an amount of time elapsed since the item became available for acquisition, or a shelf life associated with the item, as taught by Weaver, into the method of Hsu in view of Krishnan Gorumkonda would have been recognized by those of ordinary skill in the art as resulting in an improved method that would provide product transparency to buyers (Weaver, abstract).

Claims 15 and 17 recite limitations directed towards a computer program product. The limitations recited in claims 15 and 17 are parallel in nature to those addressed above for claims 5 and 7, respectively, and are therefore rejected for those same reasons set forth above in claims 7 and 7, respectively.

Claims 8 and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Hsu, in view of Krishnan Gorumkonda, in further view of Khatri et al. (US 2016/0063065 A1), hereinafter Khatri.
Regarding claim 8, Hsu in view of Krishnan Gorumkonda teaches the method of claim 1. Hsu further discloses:
-wherein generating the user interface comprising the information describing the one or more items comprises:
-responsive to determining that the satisfaction score for the item is at least the threshold ranking, generating the user interface comprising the information describing the one or more items and information indicating the of the item (Hsu, see at least: “The choice as to whether to focus on making an accurate ranking prediction or a rating prediction may ultimately depend on product considerations (i.e., how the prediction is used with respect to a product). For example, if one cares about showing the top 10 recommended products [i.e. responsive to determining that the satisfaction score for the item is at least the threshold ranking] in different verticals, then evaluating the error of the predicted rank may make more sense” [0157] and Fig. 10 shows the interface [i.e. generating the user interface comprising] of the searchable webpage includes a results webpage for the search query of ‘cameras’ and the generated predictive ratings associated with each search result [i.e. comprising the information describing the one or more items and information indicating the of the item]).

Hsu in view of Krishnan Gorumkonda does not explicitly teach generating the user interface comprising the information describing the one or more items comprises: determining that the freshness satisfaction score for the item is at least a threshold score; and responsive to determining that the freshness satisfaction score for the item is at least the threshold score, generating the user interface comprising the information describing the one or more items and the satisfaction score being a freshness satisfaction score.
Khatri, however, teaches generating a quality score of items for sale (i.e. abstract), including the known technique of generating the user interface comprising the information describing the one or more items comprises: determining that the freshness satisfaction score for the item is at least a threshold score (Khatri, see at least: “The pre-filtered item listing(s) can be filtered based on one or more of the quality score, the user attributes, and the contextual information. If the pre-filtered item listings are filtered based on user attributes only, then the quality score module 238 and the ranking module 304 can be used to filter the pre-filtered item listings based on the quality score and the recommendation module 240 can filter the pre-filtered item listings based on the quality score. If the pre-filtered item listings are pre-filtered based on the quality score, then the recommendation module 240 can further filter the pre-filtered items based on the contextual information and the user attributes if the user attributes are available. Pre-filtering based on quality score can include using the ranking module 304 and the quality score module 238 to determine a quality score and rank the item listing(s) based on the quality score. Pre-filtering based on the quality score can include comparing the quality score to a threshold and removing an item listing that is less than the threshold [i.e. determining that the freshness satisfaction score for the item is at least a threshold score]” [0056] and “As used herein “recommending” or “recommendation” mean displaying an item listing to a user [i.e. wherein generating the user interface comprising the information describing the one or more items], such as under a heading indicating that the user might be interested in the item listing” [0019] and “The quality score module 238 can determine a quality score of an item listing, such an item listing indexed in the database 126. The quality score module 238 can consider an item listing freshness (i.e. how recently the item listing was put on sale or posted for sale on the network), an item listing quality (i.e. a condition of the item listing, such as new, like new, good, slightly used, used, poor, etc.) [i.e. freshness satisfaction score]” [0050]); and
the known technique of responsive to determining that the freshness satisfaction score for the item is at least the threshold score, generating the user interface comprising the information describing the one or more items and the satisfaction score being a freshness satisfaction score (Khatri, see at least: “The pre-filtered item listing(s) can be filtered based on one or more of the quality score, the user attributes, and the contextual information. If the pre-filtered item listings are filtered based on user attributes only, then the quality score module 238 and the ranking module 304 can be used to filter the pre-filtered item listings based on the quality score and the recommendation module 240 can filter the pre-filtered item listings based on the quality score. If the pre-filtered item listings are pre-filtered based on the quality score, then the recommendation module 240 can further filter the pre-filtered items based on the contextual information and the user attributes if the user attributes are available. Pre-filtering based on quality score can include using the ranking module 304 and the quality score module 238 to determine a quality score and rank the item listing(s) based on the quality score. Pre-filtering based on the quality score can include comparing the quality score to a threshold and removing an item listing that is less than the threshold [i.e. responsive to determining that the freshness satisfaction score for the item is at least the threshold score]” [0056] and “As used herein “recommending” or “recommendation” mean displaying an item listing to a user [i.e. generating the user interface comprising the information describing the one or more items], such as under a heading indicating that the user might be interested in the item listing” [0019] and “The quality score module 238 can determine a quality score of an item listing, such an item listing indexed in the database 126. The quality score module 238 can consider an item listing freshness (i.e. how recently the item listing was put on sale or posted for sale on the network), an item listing quality (i.e. a condition of the item listing, such as new, like new, good, slightly used, used, poor, etc.) [i.e. freshness satisfaction score]” [0050]). These known techniques are applicable to the method of Hsu in view of Krishnan Gorumkonda as they both share characteristics and capabilities, namely, they are directed to generating a quality score of items for sale.
It would have been recognized that applying the known techniques of generating the user interface comprising the information describing the one or more items comprises: determining that the freshness satisfaction score for the item is at least a threshold score; and responsive to determining that the freshness satisfaction score for the item is at least the threshold score, generating the user interface comprising the information describing the one or more items and the satisfaction score being a freshness satisfaction score, as taught by Khatri, to the teachings of Hsu in view of Krishnan Gorumkonda would have yielded predictable results because the level of ordinary skill in the art demonstrated by the references applied shows the ability to incorporate such references into similar methods. Further, adding the modifications of generating the user interface comprising the information describing the one or more items comprises: determining that the freshness satisfaction score for the item is at least a threshold score; and responsive to determining that the freshness satisfaction score for the item is at least the threshold score, generating the user interface comprising the information describing the one or more items and the satisfaction score being a freshness satisfaction score, as taught by Khatri, into the method of Hsu in view of Krishnan Gorumkonda would have been recognized by those of ordinary skill in the art as resulting in an improved method that would provide relevant product recommendations to a user (Khatri, [0081]).

Claim 18 recites limitations directed towards a computer program product. The limitations recited in claim 18 are parallel in nature to those addressed above for claim 8, and are therefore rejected for those same reasons set forth above in claim 8.

Response to Arguments
Rejections under 35 U.S.C. §101
Applicant argues that the present claims do not recite a judicial exception under Step 2A Prong 1 of the revised Alice/Mayo eligibility test. The Office Action groups the present claims within the "Certain Methods of Organizing Human Activity" grouping of abstract ideas. In particular, the Office Action says that the claims are abstract because they recite the concept of "determining freshness satisfaction scores of items for purchase and providing the item data to a user." However, the amended claims are not directed to this abstract idea but rather, they recite a specific technical architecture that overcomes two distinct technical limitations inherent in online concierge systems that handle perishable inventory. The first technical problem is data drift and signal degradation. As the specification explains, "the freshness of items included in an order, such as fruits, vegetables, or baked goods, may diminish over time, making them less appealing," and "the inability of online concierge systems to provide descriptions of items that reflect the freshness of the items available at retailer locations may negatively affect the ordering experiences of their users." Filed Specification at paragraphs 1-2. Because perishable items physically decay between the time they arrive at a retailer location and the time a user requests information about them, any static or pre-computed freshness representation becomes stale. That is, the underlying data available for analysis by an online computing system drifts as the perishable item degrades. The claimed invention addresses this technical problem by providing both item lifecycle state and environmental conditions as joint input signals to a model that was previously trained on those same paired input categories. This architectural pairing is significant because freshness degradation is a function of both the item's inherent lifecycle stage (e.g., time since harvest) and extrinsic environmental factors (e.g., temperature, humidity, light exposure at the storage location). A model trained on either category alone would fail to capture their interaction. The claim addresses this by requiring that both freshness lifecycle information and current environmental information be retrieved at request time and fed to the model whose training regime explicitly included both categories of temporal inputs. This is not generic data retrieval but a specific technical integration that equips the model to account for how items degrade over time under varying environmental conditions (Remarks, pages 13-14).
Examiner respectfully disagrees. Initially, Examiner points out that Under Prong 1 it is determined whether the claim recites a judicial exception, it isn’t until Prong 2 that it is determined whether the claim recites additional elements that integrate the exception into a practical application of the exception. The claimed limitations recite concept of determining freshness satisfaction scores of items for purchase and providing the item data to a user which falls within the “Certain Methods of Organizing Human Activity” groupings of abstract ideas, enumerated in MPEP 2106.04(a) as the claims encompass advertising, and marketing or sales activities.
Even under Prong 2, the additional elements, such as the trained machine-learning model, are recited at a high level of generality (i.e. as generic computing hardware) such that they amount to nothing more than mere instructions to implement or apply the abstract idea on a generic computing hardware (or, merely use a computer as a tool to perform an abstract idea). Diminished freshness being unappealing and an inaccurate reflection of freshness negatively affecting the ordering experiences of their users are not improvements in the functioning of a computer or an improvement to another technology or technical field. Improving the accuracy of data improves the data, it does not improve technology such as the machining learning model itself. Accordingly, the claims are directed to an abstract idea and are not integrated into a practical application.

Applicant further argues that a second technical problem addressed by the claimed invention is cold-start inference for user-specific predictions. In machine learning systems that personalize outputs, a known technical challenge is inferring user preferences when a system lacks explicit preference declarations from the user at the time of the request. The claimed invention solves this cold-start problem in the context of user-specific preferences for perishable items at different stages of degradation by defining a dual-input training and inference architecture. During training, the model's parameters are adjusted based on four coordinated input categories: (1) item data including freshness lifecycle and environmental information, (2) user data for a plurality of users including historical conversion information, (3) conversion data for a plurality of conversions by those users, and (4) per-conversion satisfaction labels. At inference time, the claim requires the model to receive both "the set of item data for the item and a set of user data for the user," and specifies that "the machine-learning model derives freshness preferences of the user from the historical conversion information in the set of user data for the user." This dual-input architecture at both training and inference time enables the model to learn a mapping from historical behavioral signals to current freshness preferences, thereby producing a user-specific prediction without requiring explicit real-time preference input from the user. Together, these elements define a particular prediction architecture that fuses real-time item conditions (subject to temporal drift) with learned user behavioral representations to produce a personalized output at the moment of request. This is not a recitation of the general idea of scoring and displaying freshness information, but a specific technical pipeline: the claim requires retrieval of current lifecycle and environmental data responsive to the request, a model trained on the pairing of those same input categories with user behavioral history, and inference that takes both item data and user data as inputs to derive user-specific freshness preferences. Therefore, the present claims are not directed to an abstract idea but rather recite a technological process that overcomes specific technical limitations of data drift and cold-start personalization through a defined multi-input prediction architecture (Remarks, pages 14-15).
Examiner respectfully disagrees. Personalizing data is not a technical problem. Additionally, providing the duel inputs (i.e. specific types of data) improves the data provided, it does not improve the machine learning itself, nor does it improve the underlying computer technology. Regardless of when the data is provided it is still data and not an additional element. Accordingly, the claims are directed to an abstract idea and are not integrated into a practical application.

Applicant further argues that even if the present claims were directed to a judicial exception, that judicial exception would be integrated into a practical application under Step 2A, Prong 2 of the revised Alice/Mayo eligibility test. First, the claim requires retrieval of current lifecycle and environmental data responsive to the user's request. This is not generic data gathering but a specific architectural requirement for counteracting data drift by ensuring that the model's inputs reflect the item's present physical state rather than a stale cached value. Second, the claim defines a particular multi-input training architecture requiring four distinct categories of training data (item data with freshness lifecycle and environmental information, user data with historical conversion information, conversion data, and per-conversion satisfaction labels) and a
corresponding multi-input inference architecture that takes both item data and user data as inputs. This dual-input architecture at both training and inference time is what enables the system to derive user freshness preferences from behavioral history without explicit user input, making it a specific technical mechanism, not a mere instruction to "apply" an abstract idea on a computer. The claimed user interface is a practical application of this model output because the user-specific freshness satisfaction score is incorporated into a generated user interface as a visual indication for the user's predicted satisfaction with the current state of the perishable item, providing a concrete, real-time, personalized output that makes real-time user interactions with the interface faster and more customized. The claims therefore reflect an improvement in the functioning of the online concierge system itself by addressing the technical problems of data drift and implicit preference derivation through a defined prediction architecture rather than through generic computing. In view of the above, this rejection should be withdrawn (Remarks, pages 15-16).
Examiner respectfully disagrees. “Data drift” is not a technological problem. Additionally, deriving user freshness preferences from behavioral history without explicit user input is not a technological improvement and providing the duel inputs (i.e. specific types of data) improves the data provided, it does not improve the machine learning itself, nor does it improve the underlying computer technology. Regardless of when the data is provided it is still data and not an additional element. Furthermore, merely utilizing an interface to display the data does not improve the interface technology itself and providing a concrete, real-time, personalized output is not a technological improvement. Accordingly, the claims are ineligible.

Rejections under 35 U.S.C. §103
Applicant argues that the cited references fail to teach or disclose these claimed features. The Office Action cites Krishnan Gorumkonda as teaching user-specific desired ripeness based on user preferences. In particular, the Office Action interprets the ripeness score of a particular produce item in Krishnan Gorumkonda as being equivalent to the claimed user-specific satisfaction score because the user in Krishnan Gorumkonda indicates a desired ripeness. However, the Office Action has improperly conflated Krishnan Gorumkonda's objective ripeness regression with the claimed prediction of user-specific satisfaction with a given item. Krishnan Gorumkonda describes a "system for predicting produce quality," not a system that predicts user satisfaction. The system of Krishnan Gorumkonda observes a particular produce item with visual or infrared sensors and compares the sensed characteristics of that item in that moment against known characteristics of that category of item at different stages of ripeness along a ripeness regression. Each produce item is assigned a score along that timeline which specifies how raw or rotten the produce item is. See Krishnan Gorumkonda at col. 2 Lns. 1-18. When a customer requests an item, the request explicitly includes a desired ripeness level. Id. The Krishnan Gorumkonda system simply finds a produce item in inventory that matches an explicitly stated preference. This is fundamentally different from the claimed invention, in which the machine-learning model itself derives the user's freshness preferences from the user's historical conversion information, without requiring the user to explicitly state a preference at the time of the request (Remarks, pages 16-18).
Examiner respectfully disagrees. Hsu discloses that training data used to train a predictive machine learning model, that predicts a rating score of a product, includes historical product rating, product data, and extracted sentiments from product reviews [i.e. the machine-learning model derives preferences of the user from the historical conversion information in the set of user data] (see Hsu, [0167]-[0168], [0070], and [0178]). Krishnan modifies the derivation to be of freshness preferences and modifies the set of user data to be a set of user data for a user. Specifically, Krishnan teaches that the desired ripeness score is determined from passive communications of a user such as shopping/order history and that a user can desire a different level of ripeness depending on an intended use (see Krishnan, Col. 8 Ln. 30-39, Col. 10 Ln. 38-41, and Col. 17 Ln. 5-10). The ripeness score is a user-specific freshness satisfaction score as the desired ripeness is based on a preference of a user that is derived from passive communications of a user such as shopping/order history [i.e. deriving the user's freshness preferences from the user's historical conversion information, without requiring the user to explicitly state a preference at the time of the request]. Accordingly, the cited references teach the amended claims.

Applicant further argues that the amended claims clarify this distinction by requiring a defined training architecture in which the model's parameters are adjusted based on four distinct categories of training inputs: (1) item data including freshness lifecycle and environmental information, (2) user data for a plurality of users including historical conversion information, (3) conversion data for a plurality of conversions, and (4) per-conversion satisfaction labels. At inference time, the amended claim further requires the model to predict the user-specific freshness satisfaction score based on both the set of item data for the item and a set of user data for the requesting user and specifies that the model derives that user's freshness preferences from the historical conversion information in the user's data. Neither Krishnan Gorumkonda's sensor-based ripeness regression nor Hsu' s generic product rating prediction, alone or in combination, teaches or suggests a model with this training structure or this inference-time use of user-specific behavioral data to derive likelihood of user-specific satisfaction (Remarks, pages 18-19).
Examiner respectfully disagrees. Hsu discloses that training data used to train the parameters of a Bayesian network predictive machine learning model, that predicts a rating score of a product, includes historical product rating [i.e. the user data for the plurality of users], product data [i.e. item data], and extracted sentiments from product reviews [i.e. the conversion data], as well as, that the each extracted sentiment in associated with an estimation of how positive or negative the sentiment is [i.e. the label for each conversion of the plurality of conversions] (see Hsu, [0142], [0167]-[0168], [0070], and [0178]). Krishnan modifies the item data of Hsu to include the freshness information about a freshness lifecycle of each item and the information about environmental information associated with the locations of the items (see Krishnan, Col. 2 Ln. 53-67 & Col. 3 Ln. 1, Col. 10 Ln. 10-18, and Col. 10 Ln. 49-54). Additionally, Krishnan teaches that, at the time of the request, the desired ripeness score is determined from passive communications of a user such as shopping/order history and that a user can desire a different level of ripeness depending on an intended use of the particular request (see Krishnan, Col. 8 Ln. 30-39, Col. 10 Ln. 38-41, and Col. 17 Ln. 5-10). The ripeness score is a user-specific freshness satisfaction score as the desired ripeness is based on a current preference of a user that is derived from passive communications of a user such as shopping/order history. Accordingly, the cited references teach the amended claims.

Applicant further argues that independent Claims 11 and 20 have been amended similarly to independent Claim 1, and therefore they are distinguishable over the cited references for at least the same reasons. Additionally, the claims depending from the independent claims are patentably distinguishable over the cited art at least by virtue of their dependency on the amended independent claims (Remarks, page 19).
Examiner respectfully disagrees. As detailed in response to the arguments above, claim 1 is taught by the cited references, accordingly, claims 11 and 20, as well as the dependent claims, are taught by the cited references.

Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
-Huet et al. (US 9,081,857 B1) teaches dynamically-determining relevance or ranking of content being adjusted or improved based on the freshness or seasonality of the content.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to ARIELLE E WEINER whose telephone number is [phone]. The examiner can normally be reached M-F 8:30-5:00.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Maria-Teresa (Marissa) Thein can be reached at 571-272-6764. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.

/ARIELLE E WEINER/ Primary Examiner, Art Unit 3689

Prosecution Timeline

• Apr 30, 2024: Application Filed
• Oct 01, 2025: Non-Final Rejection mailed — §101, §103
• Dec 12, 2025: Interview Requested
• Dec 19, 2025: Applicant Interview (Telephonic)
• Dec 19, 2025: Examiner Interview Summary
• Dec 22, 2025: Response Filed
• Apr 03, 2026: Final Rejection mailed — §101, §103
• Jun 05, 2026: Request for Continued Examination
• Jun 11, 2026: Response after Non-Final Action
• Jun 24, 2026: Non-Final Rejection mailed — §101, §103 (current)