US 12,488,046 B1
Workflow content generation via video analysis
Andrew Hansen, Morrisville, NC (US); Vikram Sharma, Morrisville, NC (US); Sarang Deshpande, Morrisville, NC (US); and Sriram C Thota, Morrisville, NC (US)
Assigned to Lenovo (United States) Inc., Morrisville, CA (US)
Filed by Lenovo (United States) Inc., Morrisville, NC (US)
Filed on Oct. 1, 2024, as Appl. No. 18/903,909.
Int. Cl. G06F 16/738 (2019.01); G06V 20/40 (2022.01); G06V 20/62 (2022.01); G10L 15/183 (2013.01); G10L 25/57 (2013.01)
CPC G06F 16/738 (2019.01) [G06V 20/49 (2022.01); G06V 20/635 (2022.01); G10L 15/183 (2013.01); G10L 25/57 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A device, comprising:
a processor system; and
storage accessible to the processor system and comprising instructions executable by the processor system to:
responsive to a user query, parse data related to a video to identify discrete steps that conform to the user query, the discrete steps being steps in a workflow indicated in the video;
based on identifying the discrete steps, chunk the video into different segments that are associated with different respective discrete steps;
embed the different segments in vector space for a large language model (LLM) to retrieve and present one or more of the different segments in conformance with the user query; and
execute the LLM to retrieve and present the one or more different segments in conformance with the user query.