DETAILED ACTION
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 Interpretation
The following is a quotation of 35 U.S.C. 112(f):
(f) Element in Claim for a Combination. – An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The following is a quotation of pre-AIA 35 U.S.C. 112, sixth paragraph:
An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The claims in this application are given their broadest reasonable interpretation using the plain meaning of the claim language in light of the specification as it would be understood by one of ordinary skill in the art. The broadest reasonable interpretation of a claim element (also commonly referred to as a claim limitation) is limited by the description in the specification when 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is invoked.
As explained in MPEP § 2181, subsection I, claim limitations that meet the following three-prong test will be interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph:
(A) the claim limitation uses the term “means” or “step” or a term used as a substitute for “means” that is a generic placeholder (also called a nonce term or a non-structural term having no specific structural meaning) for performing the claimed function;
(B) the term “means” or “step” or the generic placeholder is modified by functional language, typically, but not always linked by the transition word “for” (e.g., “means for”) or another linking word or phrase, such as “configured to” or “so that”; and
(C) the term “means” or “step” or the generic placeholder is not modified by sufficient structure, material, or acts for performing the claimed function.
Use of the word “means” (or “step”) in a claim with functional language creates a rebuttable presumption that the claim limitation is to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites sufficient structure, material, or acts to entirely perform the recited function.
Absence of the word “means” (or “step”) in a claim creates a rebuttable presumption that the claim limitation is not to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is not interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites function without reciting sufficient structure, material or acts to entirely perform the recited function.
Claim limitations in this application that use the word “means” (or “step”) are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action. Conversely, claim limitations in this application that do not use the word “means” (or “step”) are not being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action.
Claims 15, 17, 18, 19, and 21 recite “means for” limitations which are interpreted under 112(f).
Claim Objections
Claim 14 is objected to because of the following informalities:
Line 2 of claim 14 seems like it should end with a colon.
Appropriate correction is required.
Claim Rejections - 35 USC § 112
The following is a quotation of the first paragraph of 35 U.S.C. 112(a):
(a) IN GENERAL.—The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.
The following is a quotation of the first paragraph of pre-AIA 35 U.S.C. 112:
The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention.
Claims 1-21 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for applications subject to pre-AIA 35 U.S.C. 112, the inventor(s), at the time the application was filed, had possession of the claimed invention.
As per Claim 1 (and similarly claim 15):
The original Specification (i.e. the original Specification of Parent Application 18/812,707, hereafter original Specification, where this application is a continuation and not a continuation-in-part) does not have written description for:
activating an activated subset of specialized agents of the plurality of specialized agents corresponding to the relevant knowledge domains; retrieving retrieved information by executing one or more information retrieval operations by the activated subset of specialized agents; aggregating the retrieved information from the activated subset of specialized agents into aggregated information, the aggregated information comprising information from a plurality of knowledge domains; providing the aggregated information to one or more h-LLMs; (The original Specification’s description of “Domain Specific Agents” appears to be limited to “The NoRAG system further comprises a Domain Specific Agents module 5512: The Domain Specific Agents module 5512 comprises several domain specific agents which retrieve appropriate specialized knowledge on the query context (e.g. web search agent, stock market agent, weather data agent, IoT data, etc). It enables the NoRAG system to adapt its responses to specific domains, improving accuracy and relevance in specialized fields”, which does not specifically describe where any agents are “activated” and used to execute retrieval operations [even though this is heavily suggested because the query context is obviously not applicable to all of the agents], and also more particularly does not describe where information from a plurality of knowledge domains is aggregated and provided to one or more h-LLMs.
As per Claim 2 (and similarly claim 16):
The original Specification does not specifically describe “a database query agent configured to access structured data repositories and enterprise databases” (querying relevant information typically involves querying a database and accessing databases that include “structured data” and databases of businesses/”enterprises” but the original Specification does not specifically use the terms “database query agent”, “structured data repositories” and “enterprise databases”, and it is not clear that these features are inherent in what is described in the original Specification)
As per Claims 3-8 and 17-21:
The original Specification does not appear to have written description for any of the limitations in claims 3-8 and 17-21.
As per Claim 9:
The original Specification does not appear to have written description for
an agent orchestrator configured to select and activate an activated subset of specialized agents of the plurality of specialized agents based on a query analysis of a received query; a communication framework configured to enable inter-agent coordination and data sharing between the plurality of specialized agents; an aggregation engine configured to produce aggregated information retrieved by the activated subset of specialized agents performing one or more information retrieval operations; and an integration interface configured to provide the aggregated information to one or more LLM processing systems (see analysis in the 112[a] rejection of claim 1).
As per Claims 10-14:
The original Specification does not appear to have written description for any of the limitations in claims 10-12 and 14, and does not have written description for “a database query agent configured to access structure data repositories and enterprise databases” in Claim 13 (see 112[a] rejection of claim 2)
The dependent claims include the issues of their respective parent claims.
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 1-21 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
As per Claim 1 (and similarly claim 15):
“the relevant knowledge domains” in line 9 of claim 1 lack antecedent basis when “one or more relevant knowledge domains” in line 7 of claim 1 refers to only one relevant knowledge domain.
It is also not clear if Applicant meant to claim “analyzing the user query to identify one or more relevant knowledge domains” in line 7 of claim 1, because “each agent of the plurality of specialized agents [is] optimized for retrieving information from a respective specific knowledge domain”, and thus each agent’s domain differs from the other agents’ domains, and in order to activate a subset of specialized agents (the subset necessarily includes plural agents) of the plurality of specialized agents (that each have a respective domain that differs from the other agents’ domains) that correspond to knowledge domains determined from the user query, there logically needs to be more than one domain identified from the user query (if there is only one domain identified from the user query, then only one of the agents corresponds to that domain and the subset of agents cannot contain plural agents because the other agents would correspond to different respective domains). Additionally, if there is only one domain associated with the query, then there logically cannot be information from a plurality of knowledge domains which are retrieved, aggregated, and combined into a comprehensive response (so it seems like “analyzing the user query to identify one or more relevant knowledge domains” should be --analyzing the user query to identify relevant knowledge domains—).
“h-LLMs” in the 5th to last line of claim 1 is understood to refer to “families of large language models” but this abbreviation is not specifically defined in claim 1.
As per Claim 3 (and similarly claim 17):
It is not clear if Applicant meant to claim “at least one of” in line 2 of claim 3, because line 6 of claim 3 recites “the agent priority” which refers to “agent priority” in line 3 of claim 3, and thus it seems like the 2nd step of claim 3 cannot be performed without the first step of claim 3 (where performing the 2nd step of claim 3 without the first step of claim 3 is within the scope of “at least one of” the 4 steps of claim 3).
It is not clear if the last 2 steps of claim3 are supposed to be part of “activat[ing] the activated subset of specialized agents” because they appear to be processes that are performed after the activated subset has already been selected and activated.
As per Claim 6 (and similarly claim 20):
“the specialized agents” in line 1 of claim 6 is ambiguous (it can refer to “a plurality of specialized agents” in line 3 of claim 1 or to “an activated subset of specialized agents” in line 8 of claim 1).
As per Claim 9:
“an aggregation engine configured to produce aggregated information retrieved by the activated subset of specialized agents performing one or more information retrieval operations” in the 5th to last line to the 3rd to last line of claim 9 does not appear to be what Applicant intended to claim (and so it is not clear if Applicant meant to claim what is currently claimed). As claimed, the “aggregated information” is both “produc[ed]” by the “aggregation engine” and “retrieved by the activated subset of specialized agents performing one or more information retrieval operations”, which is strange because, in order to be retrieved, the aggregated information should already exist (such that the agents can retrieve it), and therefore it is not clear how the aggregation engine produces aggregated information that should already exist. Applicant’s intent is fairly clearly for the aggregation engine to produce aggregated information from information retrieved by the activated subset of specialized agents performing one or more information retrieval operations, but this intent is not consistent with the grammar of the claim language.
As per Claim 10:
“each specialized agent” in line 1 of claim 10 is unclear because it is not clear if this refers to “each specialized agent of the plurality of specialized agents” (probably what Applicant meant to claim) or to “each specialized agent of the activated subset of specialized agents”.
As per Claim 14:
It is not clear if Applicant meant to claim “at least one of” in line 2 of claim 14, because line 6 of claim 14 recites “the agent priority” which refers to “agent priority” in line 3 of claim 14, and thus it seems like the 2nd step of claim 14 cannot be performed without the first step of claim 14 (where performing the 2nd step of claim 14 without the first step of claim 14 is within the scope of “at least one of” the 4 steps of claim 14).
It is also not clear if the last 2 steps of claim 14 are supposed to be part of “activat[ing] the activated subset of specialized agents” because they appear to be processes that are performed after the activated subset has already been selected and activated.
The dependent claims include the issues of their respective parent claims.
Allowable Subject Matter
The following is a statement of reasons for the indication of allowable subject matter:
As per Claim(s) 1 (and similarly claim[s] 15, and consequently claim[s] 2-8 and 16-21 which depend on claim[s] 1 and 15), the prior art of record does not teach or suggest the combination of all limitations in claim(s) 1, including (i.e. in combination with the remaining limitations in claim[s] 1) A method for implementing domain-specific agent networks in large language model systems comprising: configuring a plurality of specialized agents, each agent of the plurality of specialized agents being optimized for retrieving information from a respective specific knowledge domain; receiving a user query at a query processor; analyzing the user query to identify one or more relevant knowledge domains; activating an activated subset of specialized agents of the plurality of specialized agents corresponding to the relevant knowledge domains; retrieving retrieved information by executing one or more information retrieval operations by the activated subset of specialized agents; aggregating the retrieved information from the activated subset of specialized agents into aggregated information, the aggregated information comprising information from a plurality of knowledge domains; providing the aggregated information to one or more h-LLMs; receiving a plurality of responses from the one or more h-LLMs; and generating a comprehensive response from the plurality of responses, the comprehensive response incorporating information from the plurality of knowledge domains.
As per Claim(s) 9 (and consequently claim[s] 10-14 which depend on claim[s] 9), the prior art of record does not teach or suggest the combination of all limitations in claim(s) 9, including (i.e. in combination with the remaining limitations in claim[s] 9) A domain-specific agent network system comprising: a processor; a communication device operably coupled to the processor and configured to transmit and receive digital communications; and a non-transitory computer-readable storage medium having stored thereon software that, when executed by the processor, is operable to instantiate: a plurality of specialized agents, each specialized agent of the plurality of specialized agents being configured for a specific knowledge domain; an agent orchestrator configured to select and activate an activated subset of specialized agents of the plurality of specialized agents based on a query analysis of a received query; a communication framework configured to enable inter-agent coordination and data sharing between the plurality of specialized agents; an aggregation engine configured to produce aggregated information retrieved by the activated subset of specialized agents performing one or more information retrieval operations; and an integration interface configured to provide the aggregated information to one or more LLM processing systems.
Kennewick et al. (US 2004/0044516) teaches “Once the words and phrases have been recognized by the speech recognition engine 120, the tokens and user identification is passed to the parser 118. The parser 118 examines the tokens for the questions or commands, context and criteria. The parser 118 determines a context for an utterance by applying prior probabilities or fuzzy possibilities to keyword matching, user profile 110, and dialog history. The context of a question or command determines the domain and thereby, the domain agent 156, if any, to be evoked. For example, a question with the keywords "temperature" implies a context value of weather for the question. The parser dynamically receives keyword and associated prior probability or fuzzy possibility updates from the system agent 150 or an already active domain agent 156. Based on these probabilities or possibilities the possible contexts are scored and the top one or few are used for further processing” (paragraph 152). Paragraphs 88-89 describe where multiple agents supply data and process commands/questions and query databases, and where multiple agents return results of questions, and create a response string. Paragraph 111 describes an example of a stock domain specific agent. Paragraphs 152-154 and 161-163 describe determining a most likely context/domain (i.e. singular context/domain) and evoking an appropriate agent (i.e. singular agent). Paragraph 186 describes where multiple results are returned by multiple domain agents and where a system extracts/scrapes desired information from one or more results and gathering/combining results into a “best result” and presenting “compound results”, but in this paragraph, the user’s example utterance is “please get the value of my stock portfolio” which appears to be directed to a single context/domain. Paragraph 189 describes selecting proper agents (plural agents) and submitting queries to the plural agents, but (similar to paragraph 186) does not seem to describe a user query that has multiple domains or generating a response that includes multiple domains. For claims 1 and 15, this reference does not appear to describe aggregating multi-domain information retrieved from a subset of agents which each retrieve information from a respective specific knowledge domain and then providing the aggregated information to one or more h-LLMs (the combined/best result seems to be presented directly to the user). For claim 9, this reference does not appear to describe where data is shared between agents and providing the aggregated information to one or more LLM processing systems.
11861321 teaches “A subset of the disaggregated text portions is selected at 2008. In some embodiments, the subset of the disaggregated text portions may be selected by selecting disaggregated text portions that fall below a designated size threshold. In this way, the selected subset may be combined with the document structure prompt template to determine a document structure prompt that is sufficiently small so as to be completed by a large language model without exceeding a maximum token size for the large language model” (col. 50, lines 45-53). This reference does not appear to describe where a subset of domain-specific agents is selected and where information retrieved by the subset of domain-specific agents is aggregated and provided to LLM(s).
2025/0384332 teaches “For example and as discussed above, a RAG system (e.g., RAG system 208) is a system that is used to break relevant input documents into content portions that are small enough to fit prompt size limitations associated with a generative AI model for processing queries upon. Many generative AI models, such as LLMs, are not trained on a particular library of input documents used for a particular scenario. As such, these generative AI models lack the context to process content from the particular library of input documents. Accordingly, RAG system 208 breaks content into chunks or portions (e.g., content portion 210) that are small enough to fit prompt size limitations associated with generative AI model 206. Common indexing and retrieval techniques match user queries to the most relevant content portions, and the user query and context (one or more content portions) are combined as a prompt (e.g., prompt 212) to generative AI model 206” (paragraph 13). This reference does not appear to describe where the most relevant content portions are retrieved by a subset of domain-specific agents.
2026/0018251 teaches “FIG. 45 presents a high-level schematic of an exemplary system 4500 that uses large language models 4400 and has a RAG capability. The RAG system 4500 may receive a query 4502 from a user or an automated system. Prior to submitting the query 4502 as a prompt to the large language model 4400, the RAG system 4500 may first process the query using a retrieval component 4504. The retrieval component 404 may be configured to search an external knowledge base 4506 to find relevant data 4508 (including textual and/or non-textual data) that is relevant to the query 4502 of the user. This retrieval may be performed using various techniques, such as semantic search based on embeddings, keyword matching, hybrid approaches, or other search techniques. Semantic search refers to generating embeddings using the query 4502 and comparing the query embeddings to pre-stored embeddings generated for different chunks of data in the knowledge base 4506. The embeddings for semantic search may be generated using an embedding model 4505, which may be the same embedding model used for the large language model 4400 or a different one. Semantic search may involve the use of cosine similarity, which is a measure of the similarity of the vector corresponding to the query embeddings and the respective vector corresponding to each chunk from the knowledge base using the cosine function, or may use an alternative vector similarity metric. When the query embeddings are similar enough (e.g., above a threshold similarity), the chunk of relevant data 4508 from the knowledge base may be a “hit” for the semantic search and therefore may be retrieved from the knowledge base and added to the query 4502 as additional context. Alternatively or additionally, other techniques such as keyword search may be used to retrieve relevant data 4508” (paragraph 2206) and “The one or more items of relevant data 4508 (e.g., one or more relevant data chunks from the knowledge base) may be combined with the original user query 4502 to form an augmented prompt 4510 for the large language model 4400. The system 4500 may then provide the augmented prompt 4510 to the large language model 4400 (which may be an example of the large language model 4400 4400 described in FIG. 44). The large language model 4400 may then generate a response 4512 that takes into account its internal pre-trained knowledge as well as the specific, contextual information from the relevant data 4508” (paragraph 2207). This reference does not appear to describe where the most relevant content portions are retrieved by a subset of domain-specific agents.
Double Patenting
For clarity of the record, NO Double Patenting rejections are required between this application and any Parent/Sibling applications because the claims of the Parent/Sibling applications do not teach or suggest the new matter discussed above in the 112(a) rejections of the claims of this application.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to ERIC YEN whose telephone number is (571)272-4249. The examiner can normally be reached M-F 12:00PM -8:30PM EST.
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, RICHEMOND DORVIL can be reached at (571)272-7602. 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.
EY 2/26/2026
/ERIC YEN/ Primary Examiner, Art Unit 2658