ragapp

Why we included this project

RAGapp is for teams that want question-answering over their own documents without assembling the retrieval pipeline themselves. You start one Docker container, pick the model and point it at your knowledge sources in the admin UI, and you get a chat interface for users plus a documented REST API for hooking into your existing systems. It is built on LlamaIndex, so there is no proprietary stack to commit to, and it works with hosted models from OpenAI or Gemini as well as local models via Ollama. Ready-made Docker Compose setups bundle a vector store, which makes on-premises deployment practical for teams that cannot send data to external services. One thing worth knowing before you adopt it: the container ships with no authentication layer by design, so plan to put an API gateway in front of it.