AI-Driven Software Engineering · Distributed Systems · Scalable Platforms
I design and evolve software architectures that balance business goals, engineering quality, scalability and long-term maintainability.
🇧🇷 Arquiteto de Software e Tech Lead · IA aplicada à Engenharia de Software · Catalão, GO, Brasil
Software Architect and Technical Leader with 19+ years of experience in software development, architecture and engineering leadership.
I work at the intersection of software architecture, hands-on engineering and technical strategy, helping teams make better architectural decisions, modernize platforms and build systems that can evolve safely over time.
My current focus is expanding traditional software engineering practices into AI-driven and agentic architectures, combining proven engineering fundamentals with LLMs, RAG, AI Agents and the Model Context Protocol.
Most of my work lives in private repositories for the companies I serve. The projects below are the public side of it.
| Project | What it shows |
|---|---|
| fluxo-nfs-e-automatico-brasil | AI reads NFS-e requests from e-mail/WhatsApp; deterministic rules decide whether to issue, ask or escalate. LLM as a reader, never as the decision-maker. Node.js · n8n · Postgres · Anthropic · NFE.io |
| design-docs-webhooks-com-ia | From a 55-minute meeting transcript to PRD, RFC, FDD and 7 ADRs, every claim traced to a timestamp or a real file, with automated anti-hallucination checks |
| prompt-optimization-langsmith-eval | Prompt refactoring measured with 5 LLM-as-Judge metrics on LangSmith; reports both runs, not the best one |
| rag-pdf-langchain-pgvector | RAG pipeline: PDF ingestion into PostgreSQL + pgvector and grounded Q&A via LangChain LCEL |
- Enterprise & Solution Architecture
- Domain-Driven Design
- Clean & Modular Architecture
- Distributed Systems
- Event-Driven Architecture
- REST APIs & System Integrations
- Backend Engineering
- Performance & Scalability
- Software Modernization
- Engineering Governance
- Technical Leadership & Mentoring
- Architecture Decision Records
- Observability & Reliability
- AI-Driven Software Engineering
I prefer architectures that make complexity explicit and manageable.
Business Goals
↓
Architecture Decisions
↓
Domain & Application Design
↓
APIs · Events · Integrations
↓
Data · Infrastructure · Observability
↓
Continuous Evolution
Some principles I apply consistently:
- Business before technology
- Modularity before premature distribution
- Controllers should coordinate, not contain business rules
- Domain rules belong close to the domain
- Architecture must support change
- Observability is part of the architecture
- Performance should be measured, not assumed
- AI should augment engineering fundamentals, not replace them
My current research and engineering focus includes:
- Retrieval-Augmented Generation (RAG)
- AI Agents & Multi-Agent Systems
- Model Context Protocol (MCP)
- LLM gateways and provider abstraction
- Vector Search & pgvector
- Context and Memory Architectures
- AI-assisted SDLC
- Agentic Software Engineering
- AI observability and governance
The goal is not simply to add an LLM to an application, but to understand how AI becomes a first-class architectural capability.
Designing modular platforms with clear boundaries, maintainable domains, secure integrations and scalable backend architectures.
Working with APIs, asynchronous processing, events, integrations and system boundaries while avoiding unnecessary distributed complexity.
Evolving legacy applications incrementally through modularization, domain separation, observability, testing and technical governance.
Exploring architectures that combine LLMs, RAG, agents, tools, memory and model-provider abstraction with traditional engineering practices.
- MBA — Software & Solutions Architecture (XP Educação)
- MBA — Software Engineering with AI (Full Cycle, in progress)