You can find the documentation for building a RAG setup for videos using transcripts with timestamps in the LlamaIndex documentation at the following link: Multimodal RAG with VideoDB ^[1].

To continue talking to Dosu, mention @dosu.

For suppose I need to build a microsoft co pilot kind of tool which can be integrated into video calls . By using that tool we need to ask questions to get answers on what happened in the call and can I get relevant timestamps as outputs along with text answer . can u help me how to build that kind of tool using llama index and ollama ?