40 years running operations. Now I build the AI systems that actually fit them.
Executive experience across multinational manufacturing, paired with hands-on AI and automation development.
São Paulo, Brazil — working remotely, worldwide.
Get in touch
It's rarely the code. It's the fit.
Across decades working inside real operations — footwear, electronics, kitchenware, metallurgy, defense — I kept seeing the same pattern: software fails not because of bad engineering, but because the people building it never understood the data or the process underneath it. Standard software is standard. Fitting it to how a business actually runs rarely is.
A multi-location service network was losing time and accuracy to manual, duplicated data entry at reception — every unit doing it differently.
An AI-assisted intake system deployed across multiple locations, extending into WhatsApp for client communication, with a human-approval checkpoint before anything reaches the client.
Consistent process across every location, less manual entry, faster response times — without removing human oversight from anything client-facing.
More than 40 years in industrial operations — always in charge, never as an outside consultant. These are the results I've delivered, organized by the kind of problem they solve:
Eletrônica Selênium — cut material costs through supplier development and production process optimization (automation, faster changeovers); built an automated speaker assembly line manufactured in Japan; the restructuring ended in the company's sale to the Harman Group, owner of JBL. Medabil Tessenderlo — cut changeover time, opening up a market segment that had been out of reach before.
Digitel — implemented a structured product development process where none had existed before, replacing a pattern of defects and late launches; transferred technical know-how in thermoplastic materials. Sports footwear manufacturer (1,000+ employees) — structured product launch process that drastically cut time-to-market; complete operational diagnosis in 60 days.
CBC — Companhia Brasileira de Cartuchos — restructured the U.S. export line, recovering the relationship with Mossberg.
Philips do Brasil — technical training at the world's largest electronics multinational; first Brazilian designer to represent the country at the Eindhoven headquarters. Zivi Hércules — product development in the metalworking industry.
MBA in Business Process Management from Fundação Dom Cabral. Today I apply that same operational lens to building the systems myself — Next.js, Vercel, Supabase, AI integrations.
I don't just advise — I ship. These are products I designed and built end to end.
As a consultant, I look for one kind of problem: operations that need to run better — faster, cheaper, more consistent, higher quality — where the fix requires actually understanding the process before touching any technology.
My way in is always the same: diagnose the real process, redesign what's broken, then apply AI wherever it earns its place — never AI for its own sake.
- The same information typed into two or three different systems, by two or three different people.
- Service quality that depends on which shift or which employee is doing the work.
- One person holding too much operational knowledge in their head — a risk if they ever leave.
- Numbers that actually matter being harder to pull out of your own systems than they should be.
If any of that sounds familiar, that's the conversation I want to have.