About
Operator since 1999. Engineering discipline. Long-term support.
The 1999 story
We have been running our own businesses since 1999. Not as advisors. As operators.
We started small. We built our own systems as we grew — stock control, automated estimating, accounting integrations, customer workflows, advertising tools — the kind of operational backbone that turns a young business into one that scales without falling over. We solved problems as they arose, then built systems to anticipate the next ones before they could happen.
Python since 2020. Shipping AI tools since 2025. Twenty-seven years of building, breaking, and rebuilding the kind of operational systems we now build for clients — under our own pressure, with our own money on the line.
We did not arrive at AI through a slide deck.
What we believe
Business understanding before software
Every system we build starts by understanding how the business actually works — what the operators do day to day, where the friction is, what success looks like for them. The code follows the business, not the other way round.
LLM-agnostic by design
We build in modules. Each module does one job. Every interface is clean. Nothing is locked to one model or one vendor. When the model that is best today is no longer the best in twelve months — or the vendor changes their terms, or the API disappears — you can swap the component without rebuilding the system.
AI assists, never dictates
For any decision that materially affects a person, a human is in the loop by design. AI is there to make humans faster and better, not to replace human judgment. Where probabilistic systems are deployed, they sit inside evaluation regimes that surface drift early.
Integrated systems beat fragmented stacks
Most growing businesses have an integration problem disguised as an AI problem. The systems do not talk. The data disagrees. The reports cannot be trusted. We solve that first, because no integrated system works on top of a broken data foundation.
Long-term support, not build-and-abandon
We do not hand over a working system and disappear. The whole point of building something is that it keeps working — through staff changes, supplier changes, model changes, and business changes. We stay engaged for the long term, on terms that match how the business is actually using the systems.
What we don’t do
- We do not sell AI strategy decks without a delivery commitment behind them.
- We do not hand engagements off to a junior delivery team after sales. The person who scopes the work is the person who builds it.
- We do not claim certifications, accreditations, or partnerships that we do not actually hold.
- We do not present hypothetical case studies as real ones.
- We do not push AI where deterministic automation would do the job better.
- We do not prompt our way into systems. Real software, written and tested, with AI used where it earns its place.
- We do not lock you to a particular LLM or vendor. Components are replaceable by design.
Where we are
We are based in Porto and work with businesses across Europe.
The name
WildSun. Warmth. Energy. Something natural and powerful that you can rely on but do not control.
We chose it because we believe technology should feel that way: illuminating, not intimidating. Reliable, not capricious. A tool that lets your people see further and move faster — not a black box that demands your faith.