Hi everyone,
I wanted to share an open-source project I’ve been working on called Ignition RAG. Ignition RAG is a Retrieval-Augmented Generation (RAG) project built directly on Ignition. You can spin up the environment, upload documents, and start asking questions against your own data. I will likely be submitting this to the Ignition Exchange soon, but I think there is more value pulling it from GitHub since it contains all the necessary integrations in one docker stack.
GitHub Repo: GitHub - jp-torch/ignition-rag
Tech Stack:
- Ignition 8.3
- Ollama for running local AI models
- PostgreSQL with pgvector for document searching
Please feel free to submit any issues and pull requests on GitHub! A couple of features that I think would be great would be support for more document types (images, Excel, Word, etc.) and integration with OpenAI for faster responses. Ollama is great to run free local models, but it can be a bit slow on just a CPU!
Whether you’re curious about RAG, modern Ignition development practices, or just want a solid reference project, I hope this is useful.
Looking forward to hearing your thoughts and collaborating with the community.