AI & Software Engineer
Engineer with product MarTech experience at Cogna (MMS, integrations, and AI platform), evolving toward LLMs, RAG, and production automation—with a strong full-stack foundation (React, Next.js, Node.js, React Native) and software architecture.
2022→
Full-stack in production
FIAP
Undergrad + architecture graduate work
2k+
Members in the community I mentor
Since October 2022 I've been building digital products — React, Next.js, React Native, Node, APIs, and integrations. At Adekz TI I grew from intern (Oct 2022 — Jul 2023) to full-stack developer (Aug 2023 — Jul 2024), coordinating teams, aligning with clients, and using a broad stack: TypeScript, Tailwind, React Query, Firebase, Vue, and PHP.
At Cogna Educação (Jul 2024 — Present), I work as a contractor on the MarTech team across MMS, Transcriptor, and Orchestrator — I'm allocated directly at Cogna with formal employment via Jump Label.
Today the bottleneck isn't only the model: it's wiring LLMs to data, APIs, UX, and operations with cost, latency, and quality under control. That's why I deepen RAG, agents, automation, and observability as engineering — the same rigor I apply with Pub/Sub, Kafka, and GCP in day-to-day corporate delivery.
I share reading notes and studies on the blog and document cases in projects, with a verifiable stack.
FIAP education focused on software architecture, cloud, and AI foundations; more detail in the PDF and on LinkedIn.
Sep 2023 — Present
Founder, mentor, and community lead
Community with 2,000+ members: technical support, code reviews, workshops, and recruiter connections.
Top 12 in the FIAP × IBM Challenge — PrideJobs project (inclusive candidate–recruiter platform), presented at NEXT.
Medal at NEXT FIAP (Idwall partnership) — Criminal Hunters: web app for anonymous reporting with international data integration.
Winner of Hackaton MS — On-ffline project (hybrid teaching and flipped classroom).
Beyond code, I believe in sharing knowledge, connecting with community, and celebrating wins. Here are a few moments that shaped my path.
Placing on the FIAP podium with Criminal Hunter was a major milestone. It validated that well-applied technology can solve real problems and drive impact.
The project combined full-stack development, computer vision, and AI for an innovative public safety solution—and strengthened my ability to ship complex systems that blend craft, innovation, and purpose.
I had the honor to speak at Cogna’s Summit Marketing Platform alongside our MarTech team—covering the MarTech ecosystem, tools, trends, applied AI, orchestration, and why human judgment still matters in highly automated times.
We also highlighted Cursor as a key dev companion and showcased internal solutions: MOS (Martech Orchestrator System) and MMS (Martech Management System).
Kudos to every speaker—rich, innovative content with strong automation and applied AI throughout.
Joining Google’s Database Community Talk #4 was a great chance to go deeper on databases and connect with the community—sessions and tooling that broadened my view of scalable infrastructure.
Networking at events like this matters: exchanging real-world solutions and building relationships with peers facing similar challenges. It reinforces my belief that technology is built in community.
Attending AI Brasil Experience 2025—Brazil’s largest AI event—was transformative: inspiring talks, high-level technical content, and conversations with people shaping AI in Brazil.
It reinforced my conviction that applied AI—the kind that leaves the slide deck and ships to production—is the path I’m choosing. Thanks to Cogna Educação and the MarTech team for the VIP pass and the chance to live this moment of learning and innovation.
The future is AI—and it has already started.
Speed and precision—and a nod to my dog Trovão (“Thunder”): ship with pace without sacrificing quality.
Memory and resilience: systems that scale and keep learning with every release.
Recent articles on AI, engineering, and learning.
Chunking, latency, cost, and evaluation: what changes when you leave the notebook and enter corporate SLAs and messy data.
Patterns to decompose tasks, limit scope, and observe multi-step flows without turning your product into prompt theater.
Beyond generic benchmarks: how to define success for your use case, measure regression, and involve the business.