The Journey
Twenty years of technological eras. No company names — just the problems I solved.
The Origins
"When AI was a lab thing and I thought GWT would rule the world"
Started with a technical degree in Data Processing, then Computer Engineering. My first brush with AI was at a government research institute — building text mining tools for Portuguese agricultural literature to characterize plant diseases. Then I founded a tech startup in an incubator, building social platforms and mobile apps. I learned that building software is half the challenge — the business is the other half.
The Enterprise Years
"Workflow engines, digital signatures, and Python changes everything"
Dove into enterprise software: digital signature systems with ICP-Brazil cryptographic standards, workflow engines with graphical BPMN editors, multi-database support. Then Python entered my life and never left. Built corporate systems, automated deployments, created business dashboards, and in between freelanced building native Android and iOS apps. This was when I stopped being a Java developer and became a problem solver.
The Scale
"Tech leadership, optical networks, and learning Erlang by choice"
Led engineering at a startup — built an e-commerce platform from scratch and migrated legacy .NET systems to Python. Then something unexpected: optical telecom. Architected a Network Management System using protocols most devs never touch — YANG, NETCONF, SNMP. Built an SDN control plane for automatic multi-device network configuration. When the product line was acquired, I helped co-design the entire new platform. The BEAM taught me that concurrency doesn't have to be painful.
The Complexity
"NP-hard problems in production and AI before the hype"
Logistics at Brazil scale. Gig economy in its purest form — I engineered the full delivery operations ecosystem: driver pricing, real-time delivery tracking, route-driver matching, and orchestrating a workforce of outsourced and sub-outsourced couriers. Led a large-scale code migration from Django/Celery to FastAPI/Kafka with hexagonal architecture. Explored Rust and Kotlin for specific performance bottlenecks. Also built an AI-powered chatbot with NLU/NLP — before ChatGPT made everyone an AI expert.
The Age of Agents
"Full circle — AI left the lab"
Started studying AI in 2005. Twenty years later, I'm architecting intelligent agent ecosystems with LangChain and LangGraph, connecting LLMs with WhatsApp Business for inbound and outbound conversational marketing campaigns at scale. Full cycle approach — from code to deploy, infrastructure as code, CI/CD, observability, and production operations. The circle closed: from mining text in a research lab to orchestrating AI agents in production.
// SKILLS