Sunile Manjee

Create avatar-assisted voice search experience by integrating speech-to-text, semantic search, RAG and a synthesized avatar for responses.

Explore Fetch Surrounding Chunking, an emerging pattern in RAG that uses intelligent chunking and Elasticsearch vector database to optimize LLM responses. This approach balances data input to enhance the accuracy and relevance of LLM-generated answers through semantic hybrid search.