Integrate Computer Vision into Box AI
As far as i understand currently Box AI only works for assets which have a text representation. It would be great if you could add computer vision to Box AI (could be achieved by adding LLM tooling provided e.g. by GPT-4o-mini) in order of being able to extract text content from scanned images and ask questions about for the content.
The generated text content could be stored as a text representation as well which would speed up subsequent Box AI calls for the same asset.
2
votes
