Summary
I’ve spent the last several years on software that has to keep working while it changes: an ERP that a large retailer relies on every day, integrations with payment, shipping and tax providers, and the databases underneath all of it.
My work sits across the stack. On a given week that can mean designing a Spring Boot domain, migrating a large Vue codebase, tracing a slow query through a Firebird database, or wiring a queue-driven service on AWS. I like the problems that sit between those layers, because that is usually where production issues hide.
Today I lead a team of six engineers. I own the frontend architecture of the company’s ERP, deliver several of its business domains end to end, and started the AI support platform that now runs in production. I was named Developer of the Year in 2024 and again in 2025.
How I work
Understand before changing
Read the code, the data and the incident history first. Most legacy behavior exists for a reason, even when the reason is no longer documented.
Ship in small, reversible steps
Feature flags, shared databases and codemods instead of big-bang rewrites. If a change can be rolled back, it is much easier to ship.
Make systems observable
Metrics, logs and audit trails are part of the feature. You cannot fix what you cannot see, and you cannot prove a fix without a baseline.
Write it down
Decisions, investigations and trade-offs belong somewhere the next engineer can find them. This site is partly that habit made public.
Experience
May 2024 to present
Senior Software Engineer · Technical Lead · Pichau
Remote, Brazil
Large Brazilian hardware and gaming retailer. I joined as a developer and now lead a team of six engineers, owning the frontend architecture of the company ERP and its AI platform.
- Deliver several business-critical ERP domains end to end, from backend rules to the web interface.
- Lead the incremental modernization of a legacy desktop ERP into a web application, including a Vue 2 to Nuxt 3 migration.
- Conceived the AI support platform, built its proof of concept and lead the developers taking it to production.
- Work directly with business stakeholders to turn operational needs into requirements and a prioritized backlog.
- Built and operate the ERP observability stack with Prometheus, Grafana, Loki and Alertmanager.
Feb 2023 to May 2024
Full Stack Developer · Exati
Remote, Brazil
Smart-city software used by municipalities across Brazil. I worked closely with the CTO on the flagship platform.
- Contributed to the redesign of the flagship application used by municipalities.
- Led a new SEO- and performance-focused public website built with Astro.
- Implemented asynchronous batch uploads to S3 and refactored critical registration flows.
Technical focus
The thread running through my work is modernizing systems without stopping them. That shapes how I approach architecture: shared databases during transitions, read replicas to isolate reporting, explicit state machines for workflows with money or inventory attached, and migrations that are validated with evidence instead of assumed correct.
It also shapes how I use newer tools. Language models are useful in production when they sit inside deterministic workflows, with retrieval, audit trails and clear handoffs to people. They are a poor choice for the whole system.
Skills
- Frontend
- React
- TypeScript
- Vue / Nuxt
- Astro
- GraphQL
- Hydrogen
- Backend
- Java
- Spring Boot
- Python
- FastAPI
- Node.js
- Cloud & Infrastructure
- AWS
- Docker
- Linux
- CI/CD
- Data
- PostgreSQL
- Firebird
- Redis
- MariaDB
- Architecture
- APIs
- Distributed Systems
- Messaging
- Legacy Modernization
Career timeline
- 2023
Joined Exati as a full stack developer, working on smart-city software.
- 2024
Joined Pichau to work on the company ERP.
- 2024
Named Developer of the Year at Pichau.
- 2025
Named Developer of the Year at Pichau for the second time.
- 2026
Finishing a B.Sc. in Computer Engineering at UTFPR.
- Now
Technical lead of a six-person team, owning the ERP frontend architecture and the AI support platform.
Education
Expected Dec 2026
B.Sc. in Computer Engineering
Federal University of Technology, Paraná (UTFPR)
Capstone: Bagasort, a computer-vision sorting machine for oranges.
Current interests
Legacy modernization
Replacing systems the business depends on without stopping it: strangler patterns, shared databases and migrations that can be validated.
Reliable LLM systems
Putting language models behind deterministic workflows, retrieval and audit trails, so they behave like software rather than demos.
Distributed systems
Consensus, messaging and failure handling, from Raft implementations in coursework to queues and locks in production.
Engineering leadership
Code review, standards and architecture decisions that help a team ship faster without accumulating hidden risk.