Yeah yeah yeah agents and AI and LLMs. It’s one of those talks.
Coding agents, for now and obviously, are changing how we write software. How do these weird machinations fare against Elixir? In this talk, I’ll explore what makes Elixir uniquely suited to agentic workflows and where it falls short.
We’ll see why a few seconds in test suite startup time compound into minutes and hours of agent time, and a bunch more things like that. You’ll learn practical techniques for optimizing your Elixir codebase for having the best possible agentic experience (and, who knows, pleasing the agent robots in case the revolution ever comes).
A final disclaimer: I promise I’m not an AI fanatic - it’s a tool and I just try to get decent at using that tool. And I typed that em-dash back there, it wasn’t a robot.