AI Dev Meetup on Coding Agents with OpenAI and LangChain
Blog post from Koyeb
The recent AI developer meetup, which attracted over 350 signups, marked the first content-centric event since AI Engineer Paris 2025, focusing on the theme of coding agents. Speakers from Koyeb, OpenAI, and LangChain presented on topics ranging from agent-driven code generation to sandbox execution and reliability. Jen Person from Koyeb discussed her experience with a Lovable clone app, emphasizing the importance of sandboxed environments like Koyeb Sandboxes to safely experiment with AI-generated code. Marco Perini from LangChain illustrated the evolution from traditional AI agents to Deep Agents, highlighting how leveraging filesystems can enhance agent reliability and performance. VB Srivastav from OpenAI introduced Codex, demonstrating its capability to accelerate coding workflows by generating, completing, and refactoring code, while also addressing best practices for effective prompting. The event underscored the significance of community engagement and collaboration in advancing AI development practices.