A chatbot that answers from your own PDFs, docs and help pages, and shows where each answer came from. I use hybrid search (keyword plus vector) with reranking, so it finds exact terms like product codes as well as meaning.
What you get: document ingestion, a FastAPI question-answering endpoint with cited answers, Docker setup, and a small test set with retrieval scores so you can see how well it works instead of taking my word for it.
This is the same approach as my open-source RAG Engine project, and I co-built a Vertex AI and LlamaIndex RAG feature in a live product at my day job. Works with Gemini or OpenAI.
A chatbot that answers from your own PDFs, docs and help pages, and shows where each answer came from. I use hybrid search (keyword plus vector) with reranking, so it finds exact terms like product codes as well as meaning.
What you get: document ingestion, a FastAPI question-answering endpoint with cited answers, Docker setup, and a small test set with retrieval scores so you can see how well it works instead of taking my word for it.
This is the same approach as my open-source RAG Engine project, and I co-built a Vertex AI and LlamaIndex RAG feature in a live product at my day job. Works with Gemini or OpenAI.