Skip to content
View filipeestacio's full-sized avatar

Highlights

  • Pro

Block or report filipeestacio

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
filipeestacio/README.md

Hi there 👋

I help people put AI to work — safely

My name is Filipe Estácio. I'm a software engineer and cloud architect based in Caldas da Rainha, Portugal 🇵🇹, and I build AI systems that actually do real work — not flashy demos, but reliable helpers you can trust with the day-to-day.

Most people using AI today are chatting with it one message at a time. I go a step further: I build teams of AI agents that pick up tasks, do the work, and hand it back for a human to approve — with guardrails that stop them from ever going off the rails.

🤖 What that looks like in practice

I run my own fleet of AI agents that work together like a small, tireless team — and I built the whole system underneath them. In plain terms, that means I know how to:

  • Turn AI into a real workforce — multiple agents that plan work, write code, review each other, and tee it up for a human's yes/no. You stay in charge; they do the heavy lifting.
  • Keep it safe by design — every agent runs in its own locked-down space with strict limits. The worst a misbehaving agent can do is suggest a change you still have to approve. Nothing reaches customers without a human saying go.
  • Make AI trustworthy, not just impressive — automatic testing, code and security checks, and cost limits on everything, so you get the speed of AI without the nasty surprises (broken releases, runaway bills).
  • Prove it works — dashboards that show what the AI did, on what, and what it cost, so you're never flying blind.

The goal I build toward: AI that handles the routine work on its own, and a human who simply reviews the results — getting the leverage of a bigger team without the overhead.

🧑‍💼 Want help with this? (open to consulting)

I'm taking on a small number of AI consulting projects on the side. If your team is curious about AI but not sure how to use it safely and for real work — beyond the chatbot — that's exactly what I do. I can help you:

  • Find where AI will actually save you time and money (and where it won't)
  • Set it up so it's reliable, safe, and cost-controlled from day one
  • Get your team confident using it

📬 Let's talk: f.estacio@gmail.com · LinkedIn

💼 The day job(s)

  • 🛠️ Software Engineer @ Aircall — building communications software used by thousands of teams.
  • CTO & Co-Founder @ My Utility Connection — a UK startup I co-founded and built from the ground up on cloud infrastructure.

Before tech, I was a civil engineer — I even helped run London's water networks during the 2012 Olympics. That mix of real-world engineering and software is how I think about AI too: powerful, but only useful when it's dependable.

🧰 Under the hood

For the technically curious:

Languages: TypeScript · Rust · Nix · Python Cloud & infra: AWS (Lambda, DynamoDB, EventBridge, API Gateway, S3, CDK, CloudFormation) · event-driven architecture · NixOS · OpenTofu AI: Claude Code · Agent SDK · multi-agent systems · MCP · agent governance & observability Build & delivery: GitHub Actions · Turborepo · monorepos · self-hosted runners Frontend: React · Next.js · Astro · Vite

📫 Elsewhere

Popular repositories Loading

  1. filipeestacio filipeestacio Public

  2. cv cv Public

  3. blog blog Public

    Astro

  4. presentation presentation Public

    Generate self-contained HTML slideshows for Claude Code

  5. mac-dotfiles mac-dotfiles Public

    Shell