原文要点
原文完整内容
I'm deep in the planning phase for a new, mission-critical RAG application, and RAGFlow is currently my top candidate. Before I commit to deployment, I'd really love to hear from anyone who is running it in a live, production environment. Did it actually work out for you?
I'm mostly concerned about: Stress Testing: How does it hold up under real-world traffic? (Latency, stability, etc.)
Maintenance: How annoying is debugging a broken flow in production?
Overall performance: how good were the results? Were they production-ready?
The Killer Feature: What's the one part of RagFlow that truly saved your project?
Must know tips: must knows about best practices in ragflow and important features/configurations
If you have a moment, I’d seriously appreciate any quick advice or war stories you can share about your RAGFlow production journey! Thanks in advance!
原帖关键讨论
I really want to know how build high quality dataset for the assistant , it maybe the key points for RAGFlow. Does anyone know how to "cut" the paper correctly ?
aight
We use very little langchain code; it’s mostly used for LLM abstraction.
An AI human
来源：https://www.reddit.com/r/Rag/comments/1pawrzj/successful_ragflow_in_prod/
Hi there! I’m an engineer at RAGFlow. Welcome to the community!
We are currently on version 0.26.x, but we've got some really cool things cooking—including something awesome similar to an LLM Wiki. Plus, we’re planning to drop our much more powerful v1.0 in the coming months.
If you run into any issues or have ideas, feel free to open an issue or submit a PR on GitHub!