Update 27th October
Detailed Analysis mode: Introduced detailed analysis mode which ingests relevant clips to the LLM for hyper-relevant answers.
Significantly improved the hybrid search: Now uses RRF algorithm for better combined results.
Richer Metadata & video chunks: Each video now has richer metadata from the video intelligence API. We also increased the chunks from 8s to 16s and have noticed significant improvements to answer quality.
Improved RAG prompt: Clearer, more reliable answers.
Improved user experience: Significant UX updates for the demo.
Log in or sign up for Devpost to join the conversation.