Production work across IoT, web, mobile & SaaS

How should you read these case studies?

Each one follows the same shape: what the system is, what problem made it necessary, how it was built, and what changed once it shipped. The stack is listed on every entry because it is usually the fastest way to tell whether my experience matches your problem. Where a project has verified client reviews, they appear on the page itself rather than being summarised second-hand.

They are deliberately specific. A portfolio that says “enterprise solutions” tells you nothing; one that says a Traccar server was extended with a custom BCE protocol decoder for a 200-vehicle Gulf-region fleet tells you whether I have met your problem before.

What kinds of systems are here?

Ten production systems across five broad areas. If you know which one you are in, start there.

• Connected hardware and IoT — asset tracking over meshed networks, sensor telemetry, and GPS fleet platforms handling live position data.
• Healthcare and compliance — clinical workflow consolidation in a production hospital environment, and a multi-tenant SaaS built against Italy's ISTISAN 22/33 regulation.
• Full-stack web platforms — a multi-vendor marketplace, and an ongoing modernization of a live multi-site broadcast system.
• Mobile and desktop — native Android applications with payments, subscriptions and push, plus iOS interface design.
• AI integration — a catalogue-enrichment service built AI-native, with a human holding the consequential decisions.

Why are some of these anonymised?

A few clients are named; others are not, by agreement. Where a client cannot be named, the engineering is still described in full — the architecture, the constraints and the outcome are what make a case study useful, and none of that requires a logo. What you will not find is a real project quietly reworded to sound like a different one, or metrics that exist nowhere but the portfolio.