DETAILED ACTION
Notice of Pre-AIA or AIA Status
1. The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Response to Amendment
2. This Office Action is responsive to the Applicant’s amendment filed on April 14, 2026.3. Claims 1, 2, 4, 5, 9, 10, 11, 12, 13, 14, 15, 16, 18, 19, and 20 are pending, of which claims 1, and 16, and 20 are amended.4. Claims 3, 6, 7, 8, and 17 are cancelled by the applicant.
Response to Arguments
5. Applicant’s argument, see “Rejection Under 35 U.S.C. §101” filed on April 14, 2026 have been carefully considered. The claim now recites, among other things: a) receiving a user request, b) obtaining document information, c) generating a structured prompt, d) arranging category information and document information into defined prompt fields, e) providing instructions to the LLM, e) transmitting the prompt, f) obtaining the LLM output, and g) displaying the result. Accordingly, the 35 U.S.C. §101 rejection is withdrawn.
6. Applicant’s arguments with respect to the independent claims, see “II. Rejection Under 35 U.S.C. § 103” filed on April 14, 2026 have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
Claim Rejections - 35 USC § 103
7. 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.
8. 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.
9. 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.
10. Claims 1, 2, 4, 5, 9, 10, 11, 12, 13, 14, 15, 16, 18, 19, and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Hughes et al. U.S. 2016/0048544 A1 (hereinafter Hughes) in view of Manikani et al. US 2025/0078095A1 (hereinafter Manikani), further in view of Haikin et al. US 2025/0258850A1 (hereinafter Haikin)
Regarding claim 1, Hughes discloses a system comprising: a transceiver configured to obtain a user request to categorize a document via a user interface (Hughes [Figure 15] shows, e.g., “User Input Device(s) 1511”, “User Interface 1517”, “Transceiver 1574” and “Network Interface 1510”. These all work together in the controller. See also [0089] e.g., “When a user creates content in the form of a post, they must choose both which category it appears in … They start by selecting a category.”. This show that the user enters information – the user is expressly making a categorization selection), [wherein the user request comprises a list of categories to categorize the document] and [wherein each category in the list of categories comprises a category name and category description or characteristics]; obtain the user request and the list of categories from the transceiver (Hughes [0089] e.g., “… many categories are provided that users can select to give context to their post.”, see also [0090] e.g., “FIG. 1 shows the first step in content creation. The content channel … chosen provides posts with the first level of contextualization.”. see also [0031] e.g., “User-selectable category (e.g., metadata) field (e.g., “Lifestyles” or "Channels” or Tagtopics' identifiers).”. Together these teaches that: a list of predefined categories exists, and the user selects one through the GUI); obtain first document information associated with the document to be categorized responsive to obtaining the user request (Hughes [0005] e.g., “… receiving first data (e.g., metadata) relating to a first post object … and a first content object …”, see also [0112] e.g., “… a post record (e.g., “content”) can contain TagtopicsSM identifiers in either the title (typically a metadata field), the body (typically otherwise a content field), or both”. See also [Figure 12] It shows obtaining document information. “Selecting a channel will take a user to the channel browser…”.); display, via the categorizer module, the output on the user interface (Hughes [0106] e.g., “FIG. 10 shows the a category (content channel) list” [Figures 6-11] repeatedly show displaying: categories, posts, channels, and search results through the GUI). Hughes does not explicitly disclose: [wherein the user request comprises a list of categories to categorize the document] and [wherein each category in the list of categories comprises a category name and category description or characteristics] Manikani discloses wherein each category in the list of categories comprises a category name and category description or characteristics (Manikani [Figure 5]. Figure 5 illustrates a prompt template including separate fields such as “Category” “ASL Category” “Subcategory”, “Family Description”, “Supplier Entity Name”, “Commodity Description” and other structured prompt fields. Manikani teaches a structured prompt having separate fields for category information and document/entity information. Manikani teaches categories associated with descriptive information, including family descriptions, commodity descriptions, and hierarchy descriptions that characterize the categories used in the structured prompt. Under the broadest reasonable interpretation, such descriptive information constitutes the claimed category description or characteristics); generate, via a categorizer module, a prompt for a large language model (LLM) based on the user request (Manikani [0029] e.g., “The LLM agents provide prompts to an LLM using the tools and APIs. A prompt refers to a natural language utterance including at least one of an instruction(s), example(s), and an input.”. Manikani teaches generating a prompt for a large language model using LLM agents), wherein the prompt comprises the list of categories and the first document information (Manikani [0030] e.g., “The researching LLM agent (118) may be configured to populate a predefined template with the entity name and entity category, and other information about the supplier entity.”. Malkani teaches generating an LLM prompt that includes category information (“entity category”) together with document/entity information (“entity name” and other supplier information), and wherein, to generate the prompt, the processor: extracts the list of categories from the user request, wherein each category in the list of categories comprises the category name and the category description or characteristics (Manikani [0030] e.g., “The researching LLM agent (118) may be configured to populate a predefined template with the entity name and entity category, and other information about the supplier entity.” Malkani teaches extracting category information and incorporating that information into a structured prompt template before submission to the LLM); arranges the list of categories and the first document information into a plurality of fields of the prompt (Manikani [0030] e.g., “The researching LLM agent (118) may be configured to populate a predefined template with the entity name and entity category, and other information about the supplier entity”. Manikani teaches category information and document information into a predefined prompt template having multiple fields before a submission the LLM), wherein the plurality of fields comprises a category field including the list of categories and a document information field including the first document information (Manikani [Figure 5]. Figure 5 illustrates a prompt template including separate fields such as “Category” “ASL Category” “Subcategory”, “Family Description”, “Supplier Entity Name”, “Commodity Description” and other structured prompt fields. Manikani teaches a structured prompt having separate fields for category information and document/entity information), and wherein arrangement of the list of categories and the first document information into the plurality of fields structures the prompt to enable the LLM to categorize the document based on both the list of categories and the first document information (Manikani [0030] e.g., “The researching LLM agent (118) may be configured to populate a predefined template with the entity name and entity category, and other information about the supplier entity.”. Manikani teaches structuring the prompt with both category information and document/entity information so the LLM can perform the requested categorization task), and adds, in the plurality of fields, an instruction to cause the LLM (Manikani 0029] e.g., “A prompt refers to a natural language utterance including at least one of an instruction(s), example(s), and an input”. Manikani expressly teaches that an LLM prompt include one or more instructions directing the LLM to perform the requested task) to: select a category from the list of categories to categorize the document (Haikin [0077] e.g., “… the LLM starts with the current list of category names and determines if additional category names are needed…” Haikin teaches providing the LLM with a list of category names from which the categorization process proceeds, thereby teaching selection of categories from an existing category list); obtain, via the categorizer module, an output from the LLM responsive to transmitting the prompt (Manikani [0020] e.g., “LLMs may be pre-trained models that are designed to recognize text, summarize the text, and generate content using very large datasets.”, see also [0029] e.g., “The LLM agents provide prompts to an LLM using the tools and APIs. A prompt refers to a natural language utterance including at least one of an instruction(s), example(s)… he instruction(s) are specifically directed to the manner in which the LLM is to process the prompt. The examples may include one or more sample inputs and expected outputs. A prompt sent by an LLM agent that follows the same instructions and examples but with different input parameters is referred to as a parameterized prompt or templated prompt.”. Manikani teaches transmitting a prompt to an LLM for processing. Under the broadest reasonable interpretation, one of the ordinary skill in the art would understand that the LLM necessary returns a generated response after processing the prompt because the disclosed LLM-agent workflow depends upon the LLM processing the prompt to perform the requested task. Thus, Manikani teaches obtaining an output from the LLM responsive to transmitting the prompt) transmitting, via the categorizer module, the prompt to the LLM to identify a category for the document from the list of categories (Manikani [0020] e.g., “LLMs may be pre-trained models that are designed to recognize text, summarize the text, and generate content using very large datasets.”, see also [0029] e.g., “The LLM agents provide prompts to an LLM using the tools and APIs. A prompt refers to a natural language utterance including at least one of an instruction(s), example(s)… he instruction(s) are specifically directed to the manner in which the LLM is to process the prompt. The examples may include one or more sample inputs and expected outputs. A prompt sent by an LLM agent that follows the same instructions and examples but with different input parameters is referred to as a parameterized prompt or templated prompt.”. Manikani teaches transmitting a prompt to an LLM for processing. Under the broadest reasonable interpretation, one of the ordinary skill in the art would understand that the LLM necessary returns a generated response after processing the prompt because the disclosed LLM-agent workflow depends upon the LLM processing the prompt to perform the requested task. Thus, Manikani teaches obtaining an output from the LLM responsive to transmitting the prompt); and It would have been further obvious to one of the ordinary skill in the art before the effective filing date of the claimed invention to modify the Myer teaching in view of the teaching of Manikani by configuring the LLM-based prompt generation and processing system to receive user request and associated document information through a user interface and to present the resulting processing output though the user interface, as taught by Manikani. Such modification merely applies the known user interaction techniques of Hughes to the LLM-based categorization framework of Manikani., thereby enabling users to submit documents for processing and receive the resulting categorization in an interactive computing environment. The combination represents the predictable use of known user-interface techniques with known LLM prompt processioning techniques to obtain to expected result of facilitating user interaction with document categorization system. The combined teaching of Myer and Manikani does not expressly disclose wherein the user request comprises a list of categories to categorize the document; add a new category to the list of categories, or modify a category of the list of categories; However, Haikin discloses: wherein the user request comprises a list of categories to categorize the document (Haikin [0077] e.g., “… a current list of generated category names (as generated by all preceding insight batches) is used to seed the LLM when generating category names… the LLM starts with the current list of category names and determines if additional category names are needed…”. Haikin teaches providing the LLM with a list of category names used for categorization); to: select a category from the list of categories to categorize the document (Haikin [0077] e.g., “… the LLM starts with the current list of category names and determines if additional category names are needed…” Haikin teaches providing the LLM with a list of category names from which the categorization process proceeds, thereby teaching selection of categories from an existing category list); and
add a new category to the list of categories (Haikin [0077] e.g., “… the LLM may add additional categories names to the current list of category names (i.e., that were used as seeds)…” Haikin teaches adding one or more new category names to an existing list of category names based on the analyzed information), or modify a category of the list of categories (Haikin [0077] e.g., “… the LLM starts with the current list of category names and determines if additional category names are needed…or may keep the current list as it is” Haikin teaches dynamically updating the category inventory by evaluating the existing category list and determining whether it should remain unchanged or be expanded. Such management of the category inventory reasonably suggests modification of the maintained category list). It would have been further obvious to one of the ordinary skill in the art before the effective filing date of the claimed invention to modify the combined system of Hughes and Manikani in view of the teaching Haikin by utilizing the disclose category inventory management technique including maintaining an existing list of category names, and adding new category names, when appropriate. Incorporating the category management tech of Haikin into the LLM-based categorization framework of Hughes and Manikani would have enabled the categorization system to use and evolving category inventory during prompt generation and document categorization rather relying upon a fixed set of categories.
Claims 16 and 20 incorporate substantively all the limitations of claim 1 in a
Method and a non-transitory computer readable storage medium) and rejected under the rationale.
Regarding claim 2, the rejection of claim 1 is hereby incorporated by reference, Hughes, Manikani, and Haikin discloses a system, wherein the prompt is a generalized prompt in natural language (Manikani [0029] e.g., “A prompt refers to a natural language utterance including at least one of an instruction(s), example(s), and an input.”. Manikani expressly teaches that the prompt is a natural language utterance provided to the LLM).
Regarding claim 4, the rejection of claim 1 is hereby incorporated by reference, Hughes, Manikani, and Haikin discloses a system, wherein the first document information associated with the document comprises a first document path (Manikani [0029] e.g., “A prompt refers to a natural language utterance including at least one of an instruction(s)…”. Manikani expressly teaches that the prompt includes one or more instructions directing the LLM to perform the requested task. It would have been understanding by one of the ordinary skill int the art that such instructions include directing the LLM to categorize information using the supplied category information).
Regarding claim 5, the rejection of claim 1 is hereby incorporated by reference, Hughes, Manikani, and Haikin discloses a system, wherein the first document information associated with the document comprises a text string representing content of the document (Hughes [0054] e.g., “…the artifact prediction engine 222 can retrieve a set of digital artifacts… associated with the candidate development service from the digital artifact database 252 …” Hughes teaches obtaining document information (digital artifacts) associated with the document for processing. Under the broadest reasonable interpretation, the retrieved document information includes identifying information (e.g., document location or path) together with the document content used by the processing engine).
Regarding claim 9, the rejection of claim 1 is hereby incorporated by reference, Hughes, Manikani, and Haikin discloses a system, wherein the prompt further comprises second document information associated with a set of previously categorized documents (Manikani [0029] e.g., “A prompt refers to a natural language utterance including at least one of an instruction(s)…”. Manikani expressly teaches that the prompt includes one or more instructions directing the LLM to perform the requested task. It would have been understanding by one of the ordinary skill int the art that such instructions include directing the LLM to categorize information using the supplied category information).
Regarding claim 10, the rejection of claim 9 is hereby incorporated by reference, Hughes, Manikani, and Haikin discloses a system, wherein the second document information associated with each previously categorized document comprises a second document path and an associated category name (Haikin [0077] e.g., “… the LLM starts with the current list of category names and determines if additional category names are needed…” Haikin teaches providing the LLM with a list of category names from which the categorization process proceeds, thereby teaching selection of categories from an existing category list).
Regarding claim 11, the rejection of claim 1 is hereby incorporated by reference, Hughes, Manikani, and Haikin discloses a system, wherein the output comprises a category name for the document (Haikin [0077] e.g., “… the LLM starts with the current list of category names and determines if additional category names are needed…” Haikin teaches providing the LLM with a list of category names from which the categorization process proceeds, thereby teaching selection of categories from an existing category list).
Regarding claim 12, the rejection of claim 11 is hereby incorporated by reference, Hughes, Manikani, and Haikin discloses a system, wherein the output further comprises a mapping of a document identifier associated with the document with a corresponding category name (Hughes [0057] e.g., “The system 200 can be configured to generate, and display, a custom visual interface that presents the identified set of applicable authorization schemas … and indicates the corresponding data artifacts …”. Hughes teaches displaying the processing results through a user interface. Under the broadest reasonable interpretation, the displayed authorization schema and corresponding data artifacts constitute the output generated by the processing system, thereby teaching the output on the user interface)..
Regarding claim 13, the rejection of claim 1 is hereby incorporated by reference, Hughes, Manikani, and Haikin discloses a system further comprising a memory configured to store the prompt and the output (Manikani [Figure 1] Figure 1 illustrates system component including processors, memory, interfaces, database, and LLM agents configured to perform the discloses operation).
Regarding claim 14, the rejection of claim 13 is hereby incorporated by reference, Hughes, Manikani, and Haikin discloses a system, wherein the memory is further configured to store the document and the first document information, and wherein the processor is configured to obtain first document information from the memory (Manikani [Figure 1] Figure 1 illustrates system component including processors, memory, interfaces, database, and LLM agents configured to perform the discloses operation).
Regarding claim 15, the rejection of claim 1 is hereby incorporated by reference, Hughes, Manikani, and Haikin discloses a system, wherein the system is communicatively coupled with a plurality of data sources and a plurality of LLMs (Manikani [0029] e.g., ‘An LLM agent (e.g., the research LLM agent (118), the validation LLM agent (120), the consistency LLM agent (122), the consolidation LLM agent (124), the rephrasing LLM agent (126), the reassigning LLM agent (128)”).
Regarding claim 18, the rejection of claim 16 is hereby incorporated by reference, Hughes, Manikani, and Haikin discloses a method, wherein the first document information associated with the document comprises a first document path and a text string representing content of the document (Manikani [0029] e.g., “A prompt refers to a natural language utterance including at least one of an instruction(s)…”. Manikani expressly teaches that the prompt includes one or more instructions directing the LLM to perform the requested task. It would have been understanding by one of the ordinary skill int the art that such instructions include directing the LLM to categorize information using the supplied category information).
Regarding claim 19, the rejection of claim 16 is hereby incorporated by reference, Hughes, Manikani, and Haikin discloses a method, wherein the prompt further comprises second document information associated with a set of previously categorized documents, and wherein the second document information associated with each previously categorized document comprises a second document path (Manikani [0029] e.g., “A prompt refers to a natural language utterance including at least one of an instruction(s)…”. Manikani expressly teaches that the prompt includes one or more instructions directing the LLM to perform the requested task. It would have been understanding by one of the ordinary skill int the art that such instructions include directing the LLM to categorize information using the supplied category information).
Conclusion
11. THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
12. Any inquiry concerning this communication or earlier communications from the examiner should be directed to BERHANU MITIKU whose telephone number is (571)270-1983. The examiner can normally be reached Monday – Friday 8:30AM -4:00PM.
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, Ajay Bhatia can be reached at 571-272-3906. 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.
/BERHANU MITIKU/Examiner, Art Unit 2156
/VAISHALI SHAH/Primary Examiner, Art Unit 2156
Prosecution Insights