# US 12,596,739 B2 · Post-work summary generation, refinement, and update for field service calls > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,596,739 B2 - Document ID: US12596739B2 - Status: Granted - Kind: B2 - Application number: 18/830,472 - Filed: 2024-09-10 - Published: 2026-03-12 - Granted: 2026-04-07 - Earliest claimed: 2024-09-10 - Assignees: Salesforce, Inc. - Applicants: Salesforce, Inc. - Inventors: Puneet Dhaliwal; Shreyaa Sridhar; Nafeez Salma Chittoor Khader; Rajan Shah; Niv Garber; William Carpenter; Sindhusree Aita; Man Kit Yau - Web page: https://patentgrep.com/patent/US12596739B2 ## CPC - G06F 16/345: {Summarisation for human users} - G06F 16/3344: {using natural language analysis} - G06F 16/338: Presentation of query results - G06F 40/186: Templates - G06F 40/30: Semantic analysis - G06F 40/40: Processing or translation of natural language (natural language analysis G06F40/20; semantic analysis G06F40/30) - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) ## Abstract Disclosed herein are system, method, and computer program product aspects for textual post-work summary generation, refinement, and/or update. An example implementation operates by generating a first natural-language textual prompt based on a textual transcription of a first speech of a user and a first prompt template selected based on a work order. The implementation then queries a LLM with the first natural-language textual prompt to generate a textual post-work summary. The implementation then generates a second natural-language textual prompt based on a textual transcription of a second speech of the user and a second prompt template. The implementation then queries the LLM with the second natural-language textual prompt that directs the LLM to refine the textual post-work summary. In addition, the implementation updates the work order based on the refined textual post-work summary. In some aspects, the implementation may generate an updated textual post-work summary based on incorporating a textual transcription of a third speech of the user into the generated and/or the refined textual post-work summary, and then update the work order based on the updated textual post-work summary. ## Claims ### Claim 1 A method, comprising: generating, by one or more computing devices, a first natural-language textual prompt based on a textual transcription of a first speech of a user and a first prompt template selected based on a work order; querying, by the one or more computing devices, a large language model (LLM) with the first natural-language textual prompt to generate a first textual post-work summary that summarizes work performed by the user; generating, by the one or more computing devices, a second natural-language textual prompt based on a textual transcription of a second speech of the user and a second prompt template; querying, by the one or more computing devices, the LLM with the second natural-language textual prompt that directs the LLM to refine the first textual post-work summary to generate a refined first textual post-work summary; in response to the refined first textual post-work summary not being confirmed by the user to update the work order, querying, by the one or more computing devices, the LLM with a third natural-language textual prompt that directs the LLM, in a composite manner, to generate a second textual post-work summary and refine the second textual post-work summary, wherein the third natural-language textual prompt combines the first natural-language textual prompt and the second natural-language textual prompt; and updating, by the one or more computing devices, the work order based on a refined second textual post-work summary. ### Claim 2 (depends on 1) The method according to claim 1, further comprising: determining whether a database contains a custom prompt template associated with the work order; selecting from the database, as the first prompt template, the custom prompt template, based on the database containing the custom prompt template, or a default prompt template, based on the database not containing the custom prompt template. ### Claim 3 (depends on 2) The method according to claim 2, wherein the first prompt template comprises a latest custom prompt template of a plurality of custom prompt templates associated with the work order in the database. ### Claim 4 (depends on 1) The method according to claim 1, further comprising: generating an updated textual post-work summary based on incorporating a textual transcription of a third speech of the user into the refined second textual post-work summary, wherein the textual transcription comprises one or more edits from the user; and updating the work order based on the updated textual post-work summary. ### Claim 5 (depends on 4) The method according to claim 4, further comprising receiving a second user confirmation to use the updated textual post-work summary to update the work order. ### Claim 6 (depends on 1) The method according to claim 1, wherein the second prompt template comprises a default prompt template. ### Claim 7 (depends on 1) The method according to claim 1, further comprising receiving a first user confirmation to use the refined second textual post-work summary to update the work order. ### Claim 8 A system, comprising: a memory configured to store operations; and one or more processors configured to perform the operations, the operations comprising: generating a first natural-language textual prompt based on a textual transcription of a first speech of a user and a first prompt template selected based on a work order; querying a large language model (LLM) with the first natural-language textual prompt to generate a first textual post-work summary that summarizes work performed by the user; generating a second natural-language textual prompt based on a textual transcription of a second speech of the user and a second prompt template; querying the LLM with the second natural-language textual prompt that directs the LLM to refine the first textual post-work summary to generate a refined first textual post-work summary; in response to the refined first textual post-work summary not being confirmed by the user to update the work order, querying the LLM with a third natural-language textual prompt that directs the LLM, in a composite manner, to generate a second textual post-work summary and refine the second textual post-work summary, wherein the third natural-language textual prompt combines the first natural-language textual prompt and the second natural-language textual prompt; and updating the work order based on a refined second textual post-work summary. ### Claim 9 (depends on 8) The system according to claim 8, wherein the one or more processors are further configured to perform operations comprising: determining whether a database contains a custom prompt template associated with the work order; selecting from the database, as the first prompt template, the custom prompt template, based on the database containing the custom prompt template, or a default prompt template, based on the database not containing the custom prompt template. ### Claim 10 (depends on 9) The system according to claim 9, wherein the first prompt template comprises a latest custom prompt template of a plurality of custom prompt templates associated with the work order in the database. ### Claim 11 (depends on 8) The system according to claim 8, wherein the one or more processors are further configured to perform operations comprising: generating an updated textual post-work summary based on incorporating a textual transcription of a third speech of the user into the refined second textual post-work summary, wherein the textual transcription comprises one or more edits from the user; and updating the work order based on the updated textual post-work summary. ### Claim 12 (depends on 11) The system according to claim 11, wherein the one or more processors are further configured to perform operations comprising receiving a second user confirmation to use the updated textual post-work summary to update the work order. ### Claim 13 (depends on 8) The system according to claim 8, wherein the second prompt template comprises a default prompt template. ### Claim 14 (depends on 8) The system according to claim 8, wherein the one or more processors are further configured to perform operations comprising receiving a first user confirmation to use the refined second textual post-work summary to update the work order. ### Claim 15 A non-transitory computer-readable storage device having instructions stored thereon, execution of which, by one or more processing devices, causes one or more processors to perform operations comprising: generating a first natural-language textual prompt based on a textual transcription of a first speech of a user and a first prompt template selected based on a work order; querying a large language model (LLM) with the first natural-language textual prompt to generate a first textual post-work summary that summarizes work performed by the user; generating a second natural-language textual prompt based on a textual transcription of a second speech of the user and a second prompt template; querying the LLM with the second natural-language textual prompt that directs the LLM to refine the first textual post-work summary to generate a refined first textual post-work summary; in response to the refined first textual post-work summary not being confirmed by the user to update the work order, querying the LLM with a third natural-language textual prompt that directs the LLM, in a composite manner, to generate a second textual post-work summary and refine the second textual post-work summary, wherein the third natural-language textual prompt combines the first natural-language textual prompt and the second natural-language textual prompt; and updating the work order based on a refined second textual post-work summary. ### Claim 16 (depends on 15) The non-transitory computer-readable storage device according to claim 15, wherein the operations further comprise: determining whether a database contains a custom prompt template associated with the work order; selecting from the database, as the first prompt template, the custom prompt template, based on the database containing the custom prompt template, or a default prompt template, based on the database not containing the custom prompt template. ### Claim 17 (depends on 16) The non-transitory computer-readable storage device according to claim 16, wherein the first prompt template comprises a latest custom prompt template of a plurality of custom prompt templates associated with the work order in the database. ### Claim 18 (depends on 15) The non-transitory computer-readable storage device according to claim 15, wherein the operations further comprise: generating an updated textual post-work summary based on incorporating a textual transcription of a third speech of the user into the refined second textual post-work summary, wherein the textual transcription comprises one or more edits from the user; receiving a user confirmation to use the updated textual post-work summary to update the work order; and updating the work order based on the updated textual post-work summary. ### Claim 19 (depends on 15) The non-transitory computer-readable storage device according to claim 15, wherein the second prompt template comprises a default prompt template. ### Claim 20 (depends on 15) The non-transitory computer-readable storage device according to claim 15, wherein the operations further comprise receiving a user confirmation to use the refined second textual post-work summary to update the work order.