JS-Craft Newsletter #61 – Live class on LangChain and LangGraph, podcast with Luca Dellana, ergodicity, AI Agents

With that said, let's see the main links and topics of this edition:

🎓 In Jan 2025 I will run a live class on LangChain and LangGraph. It will be a practical workshop full of many coding exercises. You can see more details here and sign up for the waiting list. Plus, all the participants will receive a copy of my upcoming book on AI Agents and LangGraph!

🎙 In case you missed it, one week ago I published a new podcast episode with Luca Dellanna. Luca is the author of Ergodicity & Winning long term games, two of the best books I've recently read. We talked about building a sustainable career, managing risks, and AI.

📺 On the topic of AI engineering you may want to check these 2 talks by Hrishi Olickel LLMs - A Hackers Guide and A solid pattern to build LLM Applications! They are very good and full of practical advice. Btw, I will have Hrishi as a podcast quest during the next week.

📘 Speaking of AI I think LLM orchestration via code is much more powerful (and fun) than relying on just pure prompt engineering. This is one of the reasons I've written my book LangChain for JavaScript developers. As Harrison Chase, founder of LangChain, says: "It is all about communication!". Based on the growth chart of LangChain on NPM there are high chances that LangChain will be the next star of the JS world.

🦋 Many developer friends have switched from Twitter to BlueSky. I've decided to give it a try, and so far, I like the UI and overall vibe of the community. This is my BlueSky profile, if you want to connect.

✌️ My article on adding Human In The Loop for AI Agents using LangGraph was featured in the LangChain newsletter, and this makes me very happy! Currently, I am working on a long article about orchestrating multiple AI Agents with LangGraph. Hope to have it ready the next week.

🕊 On the same note I've been selected to be an official LangChain ambasador. I am super honored and thanks to the LangChain team!

💡 Learned this one the hard way! While in "dev mode" use smaller AI models to test your prompts. Why? Smaller models fail fast & often, and that's great! With big models, intelligence papers over mistakes, making things harder to debug. Plus smaller models are cheaper and faster, so you get a faster feedback loop.

👓 For the past few nights, I've been heads down reading Chip Huyen's new AI Engineer book. Even though it has a birds-eye perspective it is full of small useful details. I find myself often underlining paragraphs and making notes!

✍️ Had a longer talk with a friend explaining LangGraph & AI Agents and ended up writing an article about this: AI Agents and LangGraph.js

📦 If you want to build a portfolio of AI Engineer projects and want to get some JavaScript inspiration you can check out the GitHub profiles of Hrishi Olickel and Hassan El Mghari. Many nice projects over there!

🔋 Producing a podcast takes me 8–10 hours per episode, but it’s worth it. Preparing helps me learn, and interacting with inspiring guests gives me a huge energy boost. Thinking of starting your own? Reach out—I’d love to help!

And that’s a wrap for this edition of the JS-Craft newsletter! Be kind, keep coding, keep learning, and don't forget about the Christmas presents 🎄🎁.