Fermix is not another assistant. It is an Elixir-native, open-source AI agent runtime that runs on your machine, learns your taste, uses sandboxed tools, runs scheduled jobs, supports voice and app channels, and spins up parallel subagents for heavier work. Bring your own model and keys, keep control locally, and let your computer do real work.

Fermix started from a simple question: if tezra.io is supposed to be a lab run by agents, what should the agent running the lab actually look like? I did not want another cloud chatbot or another assistant that only answers questions. I wanted something that lives on my machine, runs in the background, remembers context, uses real tools, works across the apps I already use, and stays inside boundaries I control. Around the same time, I was reading José Valim’s article, “Why Elixir is the best language for AI,” and the argument clicked. Agents are not just prompts calling tools. They are long-running systems with workers, retries, failures, memory, plugins, scheduled jobs, tool calls, and parallel work happening at the same time. That sounded less like a chatbot problem and more like an Elixir problem. So I rebuilt the idea in Elixir.

The Elixir-native foundation is such a smart pick for agent runtimes, since BEAM's concurrency makes spinning up parallel subagents feel native rather than bolted on. Love that it's open source with bring-your-own-keys, no telemetry nonsense.

FERMIX

Always on. Always yours. Your personal AI agent

Always on. Always yours. Your personal AI agent