Hey everyone! I have some exciting news to share: I just joined CrewAI full-time as a Senior Software Developer! 🚀 It's crazy how life works. One day, you're making YouTube videos about a tool because you love AI and find it super helpful. The next, you're part of the team and helping build it! I'm super excited about this opportunity because AI agents are such a new field, and everyone is learning and experimenting together. I feel incredibly lucky to now dedicate 100% of my energy to this exciting field. Going forward, nothing will change about my YouTube channel or the community. I'm still committed to creating as many helpful videos as I can for you guys. Now that I'm all in on AI, I'm hoping to do even deeper dives into CrewAI, full-stack AI development, and some new AI topics that I think you'll really like. Also, I don't say this enough, but thank you guys so much for making this community awesome! I love hopping on our weekly group calls and meeting so many amazing people worldwide who are all interested in AI! If you have any questions about me going full-time at CrewAI or anything else, please drop a comment below or shoot me a DM! I'd be happy to answer! P.S. I was super lucky to join the CrewAI team on a trip to the AI Engineer World's Fair, which was an amazing experience. When I have more sleep, I'll be sharing a ton of lessons learned with you guys to help you on your AI journey! Thank you for being the best community out there! Cheers, Brandon

Which LLMs are best at creating jokes? I've noticed a lot of the non-open sourced ones are really afraid to hurt someones feelings.

My big win this week was using the brilliant Python library Instructor https://python.useinstructor.com/ . As a comparative newbie Python coder, I needed to spend a lot of effort with my LLM apps parsing the inconsistent output of LLM apps because they would not return responses in a consistent way. Pydantic is the typical gatekeeper tool to ensure output meets your requirements and errors if it does not meet this metric. The problem is that you still need to do a lot of work with the output to cleanse it to meet this requirement with Instructor this work is pretty much removed. For a quick overview see -

Hey guys! I just dropped another video for you guys where I show you how to use CrewAI with LangGraph. Connecting these 2 tools together will help you level up as a CrewAI developer and allow you to bulid more advanced crews. In the video, you'll see how we use CrewAI and LangGraph to monitor our emails and automatically create draft emails for us. You honestly have to watch the intro of the video to see this demo in action because it's mind blowing! Here's the link to the video if you want to learn more: https://youtu.be/5eYg1OcHm5k Here's a link to the source code: https://github.com/joaomdmoura/crewAI-examples/tree/main/CrewAI-LangGraph Let me know what you think of LangGraph + CrewAI when you get to try it out! P.S. LangGraph is a little more complex and code heavy so please don't get discouraged if it seems a little overwhelming at first. I tried to incorporate a ton of visuals to break things down as simple as possible. If you do have questions, please drop them below and I'll do my best to get you unstuck and moving in the right direction!

From their Discord. I'm excited for this one :) What's New? - Fixing task callback - Ability to set a specific agent as manager instead of having crew create your one - Ability to set system, prompt and response templates, so it works more reliable with opensource models (works better with smaller models) - Improving json and pydantic output (works better with smaller models) - Improving tool name recognition (works better with smaller models) - Improvements for tool usage (works better with smaller models) - Initial support to bring your own prompts - Fixing duplicating token calculator metrics - Adding couple new tools, Browserbase and Exa Search - Ability to create directory when saving as file - Updating dependencies - double check tools - Overall small documentation improvements - Smaller bug fixes (typos and such) - Fixing co-worker / coworker issues - Smaller Readme Updates