Langflow is a great platform for creating RAG pipelines. There is a lesser-known yet powerful component called “GitExtractor” which extracts files from a Github repo.
I have created RAG and data ingestion pipelines for Github repo-based question answering.
ingestion pipeline: for ingesting data from a Github repo into Astra DB vector store
RAG pipeline: for question answering based on retrieved content from Astra DB
Results: LLM was able to correctly answer my question about how to use speechlib to transcribe an audio file. speechlib is a library created by me and RAG pipeline gave context to LLM about speechlib.
You can follow instructions on following Github repo to create this: <a href="https://github.com/NavodPeiris/github_repo_rag_langflow" target="_blank" rel="noopener nofollow">https://github.com/NavodPeiris/github_repo_rag_langflow</a>
setup. import the flow. run. 🪼

The official daily.dev AI community. Led by thought leaders and AI experts. Stay tuned for insights, news and discussions. 

Langflow is a platform for creating RAG pipelines, featuring a component called 'GitExtractor' that extracts files from GitHub repositories. By creating RAG and data ingestion pipelines with Astra DB, users can perform question answering based on the content from GitHub repos. The provided example demonstrates successful question answering about the speechlib library using these pipelines.

RAG GitHub repos with Langflow