Like the focus on simplicity and not overthinking it!
I’d also address the topic of deployments further. Many people over engineer that or get blocked by the fact they don’t know operations very well. There, services like Fly.io or Heroku can help - you can even run your app for free often times.
For Streamlit, one caveat is that it requires WebSockets to run, which could be more complicated to setup in “production” than regular ol’ REST.
Hi Qiao! I’ve worked with LlamaIndex as well (which has its own supporters), but I would say LangChain has a larger community and is more established as the go-to framework for now. Furthermore, the recent investment by Sequoia Capital in LangChain underscores its significant potential and sets the stage for substantial growth.
You’re welcome! I’m glad you found the information helpful.
To be honest, I haven’t used CrewAI in my personal projects. However, from what I’ve seen and heard, CrewAI is designed to be user-friendly, particularly for teams focusing on workflow automation and productivity enhancement. In contrast, LangChain and LangGraph are more developer-oriented, catering to those building customized LLM-based applications. LangChain primarily supports Python for creating modular and versatile pipelines, while LangGraph emphasizes a structured, UI-driven platform for organizing and visualizing dependencies within LLM workflows.
Like the focus on simplicity and not overthinking it!
I’d also address the topic of deployments further. Many people over engineer that or get blocked by the fact they don’t know operations very well. There, services like Fly.io or Heroku can help - you can even run your app for free often times.
For Streamlit, one caveat is that it requires WebSockets to run, which could be more complicated to setup in “production” than regular ol’ REST.
Cheers and keep up the good work! 🙂
Thank you so much for enriching the content and providing valuable feedback about Streamlit!
I’ve heard about Fly.io and its simplicity for deployment—it’s definitely worth trying out.
Now in early 2025, is langchain still preferred, Are there any other better choices?
Hi Qiao! I’ve worked with LlamaIndex as well (which has its own supporters), but I would say LangChain has a larger community and is more established as the go-to framework for now. Furthermore, the recent investment by Sequoia Capital in LangChain underscores its significant potential and sets the stage for substantial growth.
https://www.sequoiacap.com/article/partnering-with-langchain-the-llm-application-framework/
Thanks for sharing, learned a lot. One more question, what do you think of crewai compared with Langchain/Langgtaph, I saw some people using it a lot.
You’re welcome! I’m glad you found the information helpful.
To be honest, I haven’t used CrewAI in my personal projects. However, from what I’ve seen and heard, CrewAI is designed to be user-friendly, particularly for teams focusing on workflow automation and productivity enhancement. In contrast, LangChain and LangGraph are more developer-oriented, catering to those building customized LLM-based applications. LangChain primarily supports Python for creating modular and versatile pipelines, while LangGraph emphasizes a structured, UI-driven platform for organizing and visualizing dependencies within LLM workflows.