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 .
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and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114.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 7/21/2026 has been entered.
Status of claims
Claims 1, 18 and 19 are amended. Claims 1-19 are presented for examination.
Response to Arguments
Applicant arguments filed on 6/22/2026 have been reviewed. Following is the response:
Response to Rejections under 35 U.S.C 103
Applicant argues “Burris, however, does not teach "filtering the combined communication based on a context window constraint of a generative artificial intelligence (AI) model to generate a filtered combined communication" or "providing the filtered combined communication as an input to the generative artificial intelligence (AI) model," as claimed.” Examiners agree with this point and the rejections over Burris (US 20200043087) in view of Khosla (US 20250005057) are withdrawn. However, upon further consideration a new ground(s) of rejection over Burris (US 20200043087) in view of Khosla (US 20250005057) and further in view of Chepkwony (US 20240289360) is applied.
Applicant further argues “Burris also does not teach "wherein the chat module routes the first
communication to a plurality of support modules including the first support module, and receives results from the plurality of support modules, and wherein the combined communication comprises the first communication and the results received from the plurality of support modules," as claimed. In Burris, the action module 112 "maps each message classification determined by message classification engine 111 into an action or sequence of actions to be performed by action module 112." (Burris, Paragraph [0043].) The action module may consult individual external sources such as a calendar program, a property management platform, or a database of applicable legal requirements. (Burris, Paragraphs [0045]- [0048].) However, this is not the same as a chat module that routes the first communication to a plurality of support modules and receives results from the plurality of support modules and forms a combined communication comprising the first communication and the results received from the plurality of support modules. In Burris, the action module itself performs actions, and it does not route to a plurality of support modules that each perform their own support operations and return results to be combined by the chat module. Rather, the action module sequentially consults external sources as needed to populate a template-based response. (Burris, Paragraphs [0043], [0050].)”. However, getting results from external sources like a calendar or property manager is the same as routing a query to those modules to get data. The action module needs data from these other modules, just like the chat module does. Thus, the action module is not distinct from the chat module. Burris does not show that routing and retrieval happen sequentially, nor do the claims state that routing happens in a non-sequential way.
Applicant argues “Khosla is directed to a natural language question answering service that uses an aggregator component to retrieve passages from search systems and an LLM component to generate answers. As described in Khosla, "the aggregator component 104 retrieves the passages from the search systems 124" and "may modify and supplement the question with the retrieved passages to form a prompt." (Khosla, Paragraph [0060].) The "LLM component 106 may receive the prompt from the aggregator component 104, the user context (optionally) from user context component 105, and generate one or more answers based on the prompt and the user context." (Khosla, Paragraph [0024].) A verifier component then checks for hallucination. (Khosla, Paragraphs [0025], [0068].)
Khosla, however, does not teach "filtering the combined communication based on a context window constraint of a generative artificial intelligence (AI) model to generate a filtered combined communication," as claimed. While Khosla does use similarity scoring to determine which passages are relevant, this filtering is based on relevance to the question, not based on a context window constraint of the LLM. As Khosla explains, "the aggregator component 104 may also use a similarity score to determine which passages retrieved are relevant (e.g., not out of scope)... The passages which pass a certain threshold (e.g., greater than 0.5) may be kept while passages under a certain threshold may not be kept (e.g., less than or equal to 0.5). The passages that are not kept may be deemed out of scope by the aggregator component 104." (Khosla, Paragraph [0045].) This similarity-based filtering determines whether passages are "……………. Rather, Khosla's aggregator retrieves static passages from search systems based on similarity matching, which are pre-existing documents and do not include results generated by performing support operations in response to a specific request. The concept of ensuring that the combined communication fits within the AI model's context window before providing it as input is completely absent from Khosla. Therefore, Khosla does not teach or suggest this feature of the claims...” Examiner agrees with this point and the rejections over Burris (US 20200043087) in view of Khosla ( US 20250005057) is withdrawn. However, upon further consideration a new ground(s) of rejection over Burris (US 20200043087) in view of Khosla (US 20250005057) and further in view of Chepkwony ( US 20240289360) is applied.
Applicant argues “Khosla also does not teach "wherein the chat module routes the first
communication to a plurality of support modules including the first support module, receives results from the plurality of support modules, and the combined communication comprises the first communication and the results received from the plurality of support modules," as claimed. While Khosla's aggregator component retrieves passages from multiple search systems (Khosla, Paragraphs [0022], [0033]), this is fundamentally different from a chat module that routes a communication to a plurality of support modules, receives results from those modules, and forms a combined communication. Khosla's search systems are static document repositories, not support modules that perform dynamic support operations in response to a user request. As Khosla explains, the aggregator component may "determine which of the search systems 124 the aggregator component 104 may retrieve passages from." (Khosla, Paragraph [0022].) This is passage retrieval based on similarity matching, not routing a communication to support modules that perform operations and return results. Therefore, Khosla does not teach or suggest this feature of the claims.” However, Khosla teaches routing the query to multiple search systems, gathering results from them, and building a final response. regardless of whether similarity matching is used, the concept not patentably distinct from the claimed invention.
Claim Rejections - 35 USC § 103
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.
And
KSR, 550 U.S. at 418, 82 USPQ2d at 1396. Exemplary rationales that may support a conclusion of obviousness include:
(A) Combining prior art elements according to known methods to yield predictable results;
(B) Simple substitution of one known element for another to obtain predictable results;
(C) Use of known technique to improve similar devices (methods, or products) in the same way;
(D) Applying a known technique to a known device (method, or product) ready for improvement to yield predictable results;
(E) "Obvious to try" – choosing from a finite number of identified, predictable solutions, with a reasonable expectation of success;
(F) Known work in one field of endeavor may prompt variations of it for use in either the same field or a different one based on design incentives or other market forces if the variations are predictable to one of ordinary skill in the art;
(G) Some teaching, suggestion, or motivation in the prior art that would have led one of ordinary skill to modify the prior art reference or to combine prior art reference teachings to arrive at the claimed invention.
See MPEP § 2143 for a discussion of the rationales listed above along with examples illustrating how the cited rationales may be used to support a finding of obviousness. See also MPEP § 2144 - § 2144.09 for additional guidance regarding support for obviousness determination.
Claims 1-6, 8-14, and 17-19 are rejected under 35 U.S.C. 103 as being unpatentable over Burris ( US 20200043087) in view of Khosla ( US 20250005057) and further in view of Chepkwony ( US 20240289360)
Regarding claim 1, Burris teaches a method comprising: receiving, from a client device connected to a property management software system (PMSS) ( computing device, Fig 1), a first communication indicating a request pertaining to the PMSS ( receive message from the messages 110 to the computing device, Para 0024); determining a category type associated with the request ( message classification engine classify/categorize messages, Para 0026) ; routing, using a chat module, the first communication to a first support module based on the determined category type ( categorize the message, Para 0041) , wherein the first support module is configured to perform one or more support operations associated with the category type of the request (leasing AI platform maps message into a plurality of actions -- identifies one or more actions associated with the first category, the actions pertaining to leasing the real estate unit, and at block 460, automatically executes the one or more actions, Para 0043-0048 ); performing, via the first support module, a first support operation associated with the request (action module performs action, Para 0053-0055, Fig 6-7; wherein the action module has plural support modules for e.g. legal requirement ( external database) , plural templates, calendar etc.);
receiving, at the chat module, a first machine communication from the first support module, the first machine communication comprising results of first support operation (for e.g. leasing AI takes action and action is retrieved from support module for e.g. time, property, other prospects, Para 0045)
combining, using the chat module, the first communication with the first machine communication to form a combined communication ( for e.g. calendar is combined with the reply; wherein the calendar is the first communication from the support module Para 0045, 0050, 0058-0060) ; wherein the chat module routes the first communication to a plurality of support modules including the first support module ( integration with at least one of a calendar program, a property management platform, or a database of applicable legal requirements, Para 0045; Additionally See Fig 7d) , and receives results from the plurality of support modules ( action module 112 may identify a time period when the management office of the property is open, when an agent at the property is available to perform the showing for the prospect, and optionally based on other considerations, such as whether other prospects are coming for a showing at the same time, etc., Para 0045-0047) , and wherein the combined communication comprises the first communication and the results received from the plurality of support modules ( calendar combined with other prospects) ;
and providing the natural language response to the indicated request through a user interface (UI) of the client device (provide the replies to the client, Fig 2, Fig 7a-d)
Burris does not explicitly teach providing, the combined communication as an input to a generative artificial intelligence (AI) model; obtaining an output of the generative Al model, the output comprising a natural language response generated based on the combined communication; and providing, the natural language response to the indicated request through a user interface (UI) of the client device
In the same field of endeavor Khosla teaches the filtered combined communication as an input to a generative artificial intelligence (AI) model ( the LLM component 106 receives the prompt and user context and determines an answer to the natural language question, Para 0061, Fig 3) ; [ Additionally, Khosla also teaches where the combined response is based on the chat module routes the first communication to a plurality of support modules including the first support module and receives results from the plurality of support modules ( The search systems 124 may be a plurality of search systems which can provide, but are not limited to, passages, documents, and QA pairs to the natural language question answering service 102. The natural language question answering service 102 may take that information and formulate an answer to a natural language question related to a network-based service and/or computer domain. , Para 0021, 0036) , and wherein the combined communication comprises the first communication and the results received from the plurality of support modules ( The aggregator component 104 may use a plurality of search systems 124 to retrieve passages from, and also receive QA pairs., Para 0036, Claim 8) ] ; obtaining an output of the generative Al model, the output comprising a natural language response generated based on the filtered combined communication ( At (8), the verifier component 108 determines if the answer was generated in error (e.g., hallucinated), Para 0068, Fig 3; wherein verify filters the language, Para 0025, 0057); and providing, the natural language response to the indicated request through a user interface (UI) of the client device send answer to the user, Fig 3)
It would have been obvious to POSITA having the teachings of Burris to further include the concept of Khosla before an effective date so to supplement, optimize, or otherwise modify the natural language query for better replies/results (Para 0009, Khosla)
Burris modified by Khosla does not explicitly teach filtering the combined communication based on a context window constraint of a generative artificial intelligence (AI) model to generate a filtered combined communication
However, Chepkwony teaches filtering the combined communication based on a context window constraint of a generative artificial intelligence (AI) model to generate a filtered combined communication (an option or constraint may be provided that allows the user to select a desired word and/or slide count. For instance, the word and/or slide count is used to determine a maximum word count property in the query that causes the LLM 108 to generate a response within the maximum word count, Para 0029)
It would have been obvious to POSITA having the teachings of Burris and Khosla to further include the concept of Chepkwony before effective filing date to make the system more user friendly by implementing features which are required by the user such as sentences within a particular word count.
Regarding claim 2, Burris as above in claim 1, teaches wherein determining a category type associated with the request comprises: performing semantic analysis to generate semantic data indicative of a category type of the request, wherein performing semantic analysis to generate semantic data comprises extracting natural language entity data from the first communication ( semantic parsing for the determining the classification of message, Para 0039, Fig 5)
Regarding claim 3, Burris as above in claim 2, teaches wherein determining a category type associated with the request further comprises: determining an intention of the request based on the extracted entity data ( for e.g. showing or property questions based on location, Fig 5, 7d, Para 0003) , and matching the intention with a category type based on at least one of keyword matching based on a set of previously established keyword pairings, or an output of a classifier neural network ( One algorithm for classifying messages is the heuristic approach. In the heuristic approach, a large number (e.g., hundreds) of messages, such as emails or text messages, for example, are taken and statistics are accumulated regarding what text (e.g., keywords) is used in a message of a certain category., Para 0015, Fig 5)
Regarding claim 4, Burris as above in claim 1, teaches wherein the request comprises at least one of: a request to summarize a document, a request to provide instructions, a request to send a communication ( fig 7a-d) , a request to draft a document, a request to provide a report comprising data associated with the PMSS ( Para 0044) , a request to generate a marketing description, a request to perform an action within the PMSS, a request to receive a link to a prebuilt report, or a request to generate a response to one or more questions ( questions replies, Para 0050, 0055)
Regarding claim 5, Burris as above in claim 1, teaches wherein the first support operation comprises at least one of: retrieving data from a database, or invoking an external API (action module consults with the database for e.g., Para 0045-0048; action module also gathers templates, Para 0050-0052)
Regarding claim 8, Khosla as above in claim 5, teaches wherein invoking an external API comprises mapping the request to one or more external API calls via a previously generated external API schema ( invoke api based on usage history etc., Para 0064, 0070-0072; further claim 5 only requires one of database schema or api schema and examiner interpretation is the retrieving only the database schema )
Regarding claim 6, Burris as above in claim 5, teaches wherein retrieving data from a database comprises at least one of: retrieving one or more documents from an unstructured database associated with the PMSS, retrieving structured data from a structured database associated with the PMSS (calendar or legal document or template (repository) Para 0045, 0047, 0050-0055)
Regarding claim 9, Burris as above in claim 1, teaches wherein the first machine communication is at least one of a response to one or more questions, data retrieved from a database, a follow-up question related to the request, or a confirmation that the one or more support operations have been performed ( response to a question, follow up, Para 0017, 0023, 0031, 0050-0051)
Regarding claim 10, Burris as above in claim 1, teaches wherein the response to the indicated request is at least one of a response to one or more questions, a follow-up question related to the request, or a confirmation that the one or more support operations have been performed ( questions/confirmation based on legal requirement, showing of the house etc., Para 0050-0055, Fig 6, Fig 7d)
Regarding claim 11, Burris as above in claim 1, teaches wherein the first support module is one of a plurality of support modules associated with a plurality of request category types, wherein each support module of the plurality of support modules is configured to perform support operations associated with one or more category types of the plurality of category types ( different actions based on the action modules based on the categories, Fig 7d; further action module retrieves data from database external servers, calendar etc. hence plurality of support modules)
Regarding claim 12, Burris modified by Khosla as above in claim 1, teaches routing, using the chat module, the first communication to a second support module configured to perform support operations associated with the category type of the request ( for e.g. invoking external database, Para 0045) ; performing, via the second support module, a second support operation associated with the request ( results from the external database, Para 0045) ; receiving, at the chat module, a second machine communication from the second support module associated with an output of the second support operation (action can be calendar, consulting database for legal answer, templates etc. and communicates back to the leasing platform for e.g. to display etc., Fig 7a-d) ; combining, using the chat module, the first communication with the second machine communication to form a second combined communication ( messages are combined with the information for e.g. times, etc. Para 0045); and providing, the second combined communication as a second input to the generative Al model (provide all context from the message and the communication as an input to the LLM, Fig 3, Para 0061, 0064, Khosla)
Regarding claim 13, Burris modified by Khosla as above in claim 1, teaches performing, via the first support module, a second support operation associated with the request ( different support for e.g. calendar or legal answer, Para 0043, 0045-0050) ; receiving, at the chat module, a second machine communication from the first support module associated with an output of the second support operation ( AI leasing platform receives appropriate response, Fig 7a-d); combining, using the chat module, the first communication with the second machine communication to form a second combined communication ( messages are combined with the information for e.g. times, etc. Para 0045); and providing the second combined communication as a second input to the generative Al model. (provide all context from the message and the communication as an input to the LLM, Fig 3, Para 0061, 0064, Khosla)
Regarding claim 14, Burris, modified by Khosla as above in claim 1, teaches wherein the Generative AI model has been trained on a corpus of text to create a foundation model (trained AI, Fig 3, Khosla; trained machine learning model, Para 0020, Burris)
Regarding claim 17, Khosla as above in claim 1, teaches wherein a retrieval component of a retrieval-augmented generation (RAG) system provides context associated with the request to the generative AI model (RAG techniques, Para 0029, 0079)
Regarding claim 18, arguments analogous to claim 1, are applicable. In addition, Burris teaches A system comprising: a memory device; and a processing device communicatively coupled to the memory device, wherein the processing device is to perform the steps in claim 1 (Fig 1)
Regarding claim 19, arguments analogous to claim 1 are applicable. In addition, Burris teaches non-transitory computer readable storage medium comprising instructions that, when executed by a processing device, causes the processing device to perform operations in claim 1. (Para 0025)
Claim 7 is rejected under 35 U.S.C. 103 as being unpatentable over Burris (US 20200043087) in view of Khosla (US 20250005057) and further in view of Chepkwony (US 20240289360) and further in view of Relan (US 20220374420)
Regarding claim 7, Burris as above in claim 5, teaches wherein retrieving data from a database further comprises mapping the request to a database query via a previously generated database (for example the requirement document or calendar is already available and mapping would be based on users’ message, Para 0045-0047)
While Burris does not explicitly mention mapping to database schema
Relan teaches mapping to database schema (a database is based on schema based on previous stored records etc., Para 0012, 0046)
It would have been obvious to have the teachings of Burris and Khosla to further include the concept of Relan before effective filing date to provide the most accurate response (Para 0008, Relan)
Claim 15-16 are rejected under 35 U.S.C. 103 as being unpatentable over Burris (US 20200043087) in view of Khosla (US 20250005057) and further in view of Chepkwony (US 20240289360) and further in view of Lev (US 20240370498)
Regarding claim 15, Burris as above in claim 1, mentions a custom trained AI model (Para 0056-0057) and generative model trained on specific domain (Para 0062, Khosla) Burris modified by Khosla does not explicitly teaches wherein the generative AI model has been fine-tuned on proprietary organizational data associated with property management
However, Lev teaches wherein the generative AI model has been fine-tuned on proprietary organizational data associated with property management (fine-tuned for a specific task, Para 0096; task can be real estate queries, Para 0120, 0123)
It would have been obvious having the teachings of Burris and Khosla to further include the concept of Lev before effective filing date since it’s a well-known concept of leverage gpt model for a specific task (Para 0096, 0123, Lev)
Regarding claim 16, Burris as above in claim 1, teaches AI model is fined tuned (Para 0056-0057- custom AI for real estate) and Khosla teaches generate AI (trained for e specific task, Para 0062) however does not explicitly teach wherein the generative AI model has been fine-tuned for application to PMSSs
However, Lev teaches wherein the generative AI model has been fine-tuned for application to PMSSs (fine- tuned for a specific task, Para 0096; task can be real estate queries, Para 0120, 0123)
It would have been obvious having the teachings of Burris and Khosla to further include the concept of Lev before effective filing date since it’s a well-known concept of leverage gpt model for a specific task (Para 0096, 0123, Lev)
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
US 11288322 – discusses API schema similar to claim 8
US 20230259821 – routing to different LLM model based on diagnostics.
US 20230316000 – teaches the combined communication is an input the LLM model and LLM generates a final response
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Richa Sonifrank whose telephone number is (571)272-5357. The examiner can normally be reached M-T 7AM - 5:30PM.
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, Phan Hai can be reached at (571)272-6338. 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.
/Richa Sonifrank/Primary Examiner, Art Unit 2654
Prosecution Insights