Hi everyone! I’m curious about what libraries and platforms people are using when developing AI Agents for real-world applications.
Personally, I’ve been working with langgraph for my hobby projects and experiments, but I’m wondering what others in the community are using for production systems.
If you’re not using langgraph, I’d love to hear about your preferred framework and what made you choose it over other options. Are there specific features or limitations that influenced your decision?
I’ve been using CrewAI for most of my production deployments and it’s been solid. Switched from other frameworks mainly because it handles multi-agent coordination really well without making things complicated. CrewAI’s great at managing API rate limits and error handling - both super important when you’re scaling up. The docs are decent and community support has gotten much better lately. Had some headaches early on with custom tool integration, but recent updates fixed most of that. For simple single-agent stuff, I’ll sometimes go with AutoGen since it’s easier to set up, but CrewAI’s still my go-to for anything involving multiple agents.