YogSadhna Varga

The application uses a Retrieval-Augmented Generation (RAG) architecture to combine the power of LLMs with a custom knowledge base. The heart of the system, orchestrating the retrieval of relevant documents from the vector database and augmenting the LLM prompt with this context. It includes: - **Personalized Prompts**: Uses conversation history and user profile for context. - **Intent Analysis**: Understands what the user is trying to accomplish. - **Confidence Scoring**: Measures response quality and falls back when needed. An abstraction layer that allows seamless switching between Pinecone (for production) and ChromaDB (for development). It handles semantic search to find the most relevant information for user queries. Maintains context across interactions, building user profiles and tracking conversation history to provide personalized and coherent responses.