I've spent 17 years inside the systems you're afraid to break.

Every company I talk to is having the same quiet argument with itself. One side sees what LLMs and autonomous agents can do right now — the productivity is real, the competitors are already moving. The other side has watched what happens when you bolt fast-moving technology onto a decades-old, regulated system and skip the legal questions: data leaving the building, models nobody approved, a breach nobody can explain. dezotech started in that argument.

I'm Andre Queiroz. For seventeen years I built and shipped production systems inside banking, government, and enterprise — places where "move fast and break things" is a great way to end up explaining yourself to a regulator. I learned modernization the slow, safe way: strangler-fig migrations that cut over one slice at a time, with a rollback at every step, so the business never stops while the old system quietly retires.

When LLMs and agents arrived, I watched capable teams freeze — not because the technology was hard, but because nobody could answer the audit questions. Where does the data go? Who signed off on this model? What happens when it's confidently wrong? "Shadow AI" isn't really a tooling problem. It's people shipping value faster than governance can keep up, and hoping nobody asks. I build the path that lets them stop hoping.