Hi, OpenAI Devs,
We’ve been building R2R, an open source framework for rapid development and deployment of RAG pipelines.
With R2R you can launch a basic RAG pipeline in under a minute, using the cloud platform we are building at sciphi.ai to pair with R2R. These pipelines are fully customizable, and you can readily add custom logic from LangChain or LlamaIndex or elsewhere into the template logic.
The default provider for embeddings / RAG completions is OpenAI, and the out of the box experience is similar to OpenAI Assistants with observability + evaluation built in.
Please take a look, deploy a pipeline and then let us know what features you most want added - we would really appreciate your feedback!