RAG-ChatBot-app is a chatbot web application to extract and analyze information from documents through Retrieval-Augmented Generation (RAG), built on the powerful Gradio web interface.
This project aims to create an intelligent chatbot that supports a variety of documents, allowing users to ask questions based on their own files such as PDFs, spreadsheets or texts. Using language models (LLMs), vector embeddings and a vector database, the system is able to provide contextualized and accurate answers. The application is built with a focus on local usability via Gradio and Docker, being flexible for use with different storage vectors, such as Qdrant, ChromaDB and Pinecone. The user-friendly interface developed with Gradio facilitates interaction with the chatbot in a simple way and accessible directly through the browser. The diagram below shows the communication flow between the parties involved in the RAG process performed by the application.
- Support for proprietary LLM models
- Support for local LLM models (Ollama)
- Pre-configured vector storages: Chromadb | Qdrant | Pinecone
- User authentication control
- User-friendly visual for documents analysis powered by awesome Gradio web interface
- Supports .pdf, .csv, .xls, .xlsx, txt and .docx document types
- Ready to deploy wtith Docker
- Download the repository
git clone https://github.com/fab2112/RAG-ChatBot-app.git
cd RAG-ChatBot-app
- Set the .env file with keys necessary based in custom settings
- Build docker services
docker-compose up --build -d
- After docker-compose building access the application in the browser at http://0.0.0.0:7860
- Choose your model and start a chat
- Load docs to vector database defined in settings
- The docs are split in chunks of texts + correspondent vector, and loaded to database
- This table displays all documents that have been loaded into the database.
- Table attributes: Time, File, Size, Type and Chunk-IDs
- Custom definitions for processing docs in RAG mode
- For normal chat, select RAG-mode OFF
- Access custom settings via the settings.py file for change defaut before building
| Variable | Details |
|---|---|
| USERS | Set users authentication login |
| LANGUAGE | LLM response language |
| OLLAMA_URL | Ollama internal Docker url |
| QDRANT_URL | Qdrant internal Docker url |
| CHROMADB_URL | Chroma internal Docker url |
| RETRIEVER | Database retriever |
| VECSTORAGE | Vector database |
| DATABASE_NAME | Name of space in vector database |
| PINECONE_REGION | Pinecone database on-premises infrastructure |
| PINECONE_CLOUD | Pinecone database cloud |
| EMBEDDINGS_RATELIMIT | Ratelimit of embeddings chunks per seconds |
| MODELS | LLM models |
| EMBEDDINGS_MODEL | Model embeddings and dense vector dim |
- Making the service accessible through Docker
- Add the following line under the [Service] section in "/etc/systemd/system/ollama.service"
Environment="OLLAMA_HOST=0.0.0.0"
- Save, exit, reload the systemd configuration and restart Ollama
systemctl daemon-reload
systemctl restart ollama
- Load the desired models in local host
- Search for the best model that suits you https://ollama.com/
ollama run gemma3:1b # model for chat
ollama run nomic-embed-text:latest # model for embeddings
- Infrastructure
| Component | Version |
|---|---|
| Docker Engine | 28.0.4 |
| Docker Compose | 2.34.0 |
| Ollama | 0.6.3 |
- App interface
