Banking industry
30+ production ML models and internationally award-winning work across data science and digital strategy.
About Miguel
A technical partner with real production AI experience.
I work with SMEs that want AI to improve operations in a practical way. That means fewer buzzwords, clearer decisions, and systems that hold up once they are live.
Hi, I'm Miguel Cabrita.
I have worked in data and AI for over a decade, combining mathematics, data science, product thinking, and delivery in real business environments.
Early in that path, I built machine learning systems in production in the banking industry, where the work was recognized with international awards including Banking Tech Awards, Portugal Digital Awards, and Financial Innovation Awards.
Later, as an EU-backed AI researcher, I worked with EDP Renewables North America in Houston on applied AI research focused on real operational problems in renewables, not lab experiments.
Today I bring that same level of rigor to SMEs: AI advisory, workflow design, implementation, and training that fit the realities of a small or mid-sized team. I am also a certified trainer in Portugal, holding a Certificate of Pedagogical Competencies (CCP).
I also build my own businesses, including Portablebit, which keeps me grounded in the trade-offs founders and operators actually face when choosing tools, spending money, and deciding what to automate.
Outside work, you may find me running, climbing mountains, or making music.
Experience behind the work
The common thread is practical delivery: production AI, applied research, founder-level trade-offs, and implementation plans that survive contact with real operations.
EDP Renewables NA
EU-backed applied AI research in the United States, part of an EU-USA funded research collaboration.
Two-time founder
Founder of two data and AI companies, building products and teams from the ground up in Portugal and Europe.
Client funding support
Helped secure more than EUR50K in EU innovation funding and built practical implementation plans behind it.
How I work
I work as a technical partner first: understand the business process, decide what is worth changing, then bring the right people and systems around the problem.
Start with the process
Before choosing tools, I map how work currently moves, where decisions get stuck, and which bottlenecks are actually worth automating.
Build only what will be used
I prefer clear trade-offs, small useful releases, and systems that fit the team. If something should stay manual, I will say that directly.
Bring the right specialists
When a project needs more than one specialist, I bring in vetted experts I have worked with across data engineering, backend, full stack, front end, cyber security, machine learning, AI engineering, mathematical modeling, VR/AR, and blockchain.
Stay accountable for delivery
The goal is not to add vendors around a problem. I keep the work coordinated, practical, and tied to the operational outcome we agreed on.
Build company-owned IP
Where it makes sense, I help you build systems, data assets, and operational know-how your company owns. That increases enterprise value and scalability instead of leaving critical workflows dependent on third-party software with weak data ownership.
That gives clients senior AI guidance plus access to a broader delivery bench, without turning the engagement into bloated consulting overhead.
If your business is carrying too much repetitive work, we should talk.
The first step is usually a working session to look at the current process, the bottlenecks, and what is worth automating.
Book a Session