First notes: experiments, control planes, and a bit of AI

· 2 min read

This is the first post on this small site.

The goal here is simple: to share some of the things I’m currently building, exploring, and occasionally struggling with. Most of my work these days revolves around Envoy Proxy, control planes, and the question of how we can make complex systems easier to reason about and operate.

Lately, that exploration has started to include AI and LLMs—not as a buzzword, but as a practical tool. I’m particularly interested in how they can improve developer workflows, especially in areas like:

  • configuration generation
  • debugging distributed systems
  • working with spec-driven infrastructure

A note on tools and languages

Most of my work is rooted in functional programming, especially Elixir and Erlang/OTP.

I’ve been using these languages for more than a decade, and they’ve shaped how I think about systems: concurrency, fault tolerance, and building things that keep working under pressure.

Even now, with LLMs capable of generating code in almost any language, I still find these ecosystems particularly relevant.

If anything, the rise of AI makes good abstractions more important, not less.

When code can be generated at high speed, the limiting factor becomes:

how well we can understand, validate, and operate what was generated.

Languages and platforms that provide clear, high-level abstractions make it much easier for a human to stay in the loop and reason about the system as a whole.

What this site is (and isn’t)

This isn’t meant to be a polished blog or a product showcase.

It’s more of a place to:

  • write down small ideas
  • document experiments
  • connect dots between different areas (APIs, control planes, AI, developer tooling)

Some posts will be rough. Some ideas might not go anywhere. That’s part of the process.

A small experiment in writing

This post itself is part of an experiment.

English isn’t my native language, but I still want to share ideas in a way that’s accessible. So I’m trying a workflow where I:

  • capture thoughts using audio
  • convert them to text
  • use AI to refine and structure them

The result is something that hopefully still reflects my thinking, just expressed more clearly.

What’s next

If you’re interested in:

  • Envoy and control planes
  • spec-driven infrastructure
  • AI in real-world engineering workflows

…you’ll probably find something useful here over time.

And if you have thoughts, feedback, or want to compare notes—feel free to reach out.