Lucas Garcia

I'm a backend engineer with 6 years of experience building systems where reliability isn't optional — and for the past year, my focus has been taking LLMs from demo to production. At Serasa Experian (Brazil's largest credit bureau, part of Experian plc, FTSE 100), I architected Codebuddy: an enterprise AI platform built on MCP (Model Context Protocol) that delivers LLM-powered skills — stack-specific assistants, automated PSA generation, compliance workflows — into any IDE, standardized to company policy. It's now used by 300+ engineers, with an innersource community contributing 10–15 requests per week. What I bring: — Production AI engineering: RAG pipelines, agentic workflows, and LLM orchestration integrated with enterprise Java/.NET stacks — where AI has to meet SLAs, compliance (LGPD), and real data. — Polyglot backend depth: TypeScript/Node.js (3+ years shipping enterprise services), C#/.NET Core, and Java/Spring Boot. Architecture that ages well: SOLID, Clean Architecture, DDD applied where they pay off. — Measurable impact: at Semantix, cut legacy query processing from ~10 minutes to under 1 minute for clients like Via Varejo and Colgate. Currently completing a Postgraduate degree in AI Engineering. Fluent English (daily use). São Paulo, Brazil (UTC-3) — 4–6h overlap with US and EU working hours. Open to remote full-time or contractor (B2B) roles with international teams. lucasgarcia.fg.dev@gmail.com