Sorocaba · São Paulo · Brazil · remote worldwide
Before AI, there was engineering, 12+ years building systems where software meets the physical world.
My career started long before generative AI. For over 12 years I've worked where software meets the physical world, supervisory systems, PLC-based automation on Siemens and Rockwell platforms, and electronic and hydraulic test rigs communicating over industrial networks like CAN J1939. Those projects taught me something that still shapes how I build: software only creates value when it understands the system around it. Over time, signals became data, machines became connected systems, inspection became computer vision, and AI became the next layer of the stack.
I didn't move from engineering to AI. I brought AI into engineering.
PLC programming, supervisory systems and industrial control.
Linking controllers, sensors and test rigs to their machine data.
Giving automation the ability to interpret the physical world.
Adding context, knowledge and intelligent decision support.
Connecting AI back into real engineering and business workflows.
I approach an AI system the way I approach an industrial one: understand the process, define the interfaces, validate the behavior, measure the result, and make sure it works outside the demo.
At CNH Industrial Latin America, I led AI adoption across manufacturing, quality and engineering, architecting and deploying an end-to-end inspection solution that combines CNN-based computer vision, agentic approval workflows and RAG over engineering documentation, integrated directly into the production line.
I train and evaluate LLMs where technical correctness actually matters, pairing model assessment with hands-on engineering domain expertise.
Training and evaluating LLMs across engineering domains, electrical, mechanical and industrial automation, assessing technical correctness and real-world applicability.
An end-to-end fleet monitoring SaaS I designed and built, live GPS tracking over MQTT, trajectory history with GPS replay, a fleet dashboard, restricted-zone geofencing and speed/stop alerts. Built for real operations on the plant floor.
Understand the workflow, constraints and business problem.
Identify where AI actually adds value.
Prototype the AI, agent, RAG, vision or automation workflow.
Measure model behavior, reliability and operational fit.
Integrate with existing systems and workflows.
Monitor outcomes and continuously optimize.
I'm an AI Training Consultant and LLM Trainer with 8+ years leading technical projects and putting technology into real operational environments, not just experiments.
My differentiator is the combination: I understand business processes, engineering problems, AI systems and production environments, so AI solutions actually work reliably, interact with existing systems and generate measurable results.
Practical content on AI tools, AI engineering and real-world workflows, turning hands-on experience into things teams can use.
Whether you're exploring an AI use case, building an agentic workflow, modernizing an industrial process, or looking for someone who bridges engineering and generative AI, let's talk.
