Misael Pérez Chamorro is a software architect with over 18 years of experience designing and building distributed systems, mobile apps, and backend platforms across Elixir, Erlang, Go, and Ruby. He’s an Elixir consultant who builds and ships challenging Elixir applications, with a focus on real-time systems and production-grade architecture.
He’s the host of the virtual Elixir Meetup in Mexico and a passionate advocate for using OTP to solve real-world coordination and scalability problems. When he’s not coding, you’ll probably find him on a bike ride or exploring open-world video games. Misael enjoys systems that are clean, resilient, and fun to build — which is why he keeps coming back to the BEAM.
AI is showing up in everything lately, and as an Elixir engineer it’s not always obvious where to start without falling into the usual toy demos (Is this a banana?). In this talk I’ll share a real app that runs fully locally on a Raspberry Pi: a voice-controlled smart-light system where Elixir owns the whole stack. We’ll walk through the architecture from end to end. You’ll see how Membrane captures and chunks audio, how a Phoenix LiveView dashboard makes the system easy to observe and tweak, and where Nx/Axon fits when you want small, practical models like wake-word detection and intent classification. I’ll keep it grounded in the “how”: turning a rough prototype into something usable, collecting just enough data to train a first model, and wiring the pieces together so the result feels like a real feature, not a science project. By the end, you’ll have a clear picture of where Elixir fits in AI/ML today, plus a working smart control device in your house powered by Elixir that you can keep extending as you learn.
Key Takeaways:
Target Audience: