12+ years architecting cloud and software for 40+ clients. The last year focused on agentic AI, RAG pipelines, and multi-provider LLM integrations — embedded into real client products, not demos.
I'm a Solutions Architect with 12+ years building cloud and software systems for 40+ clients across manufacturing, SaaS, healthcare, agribusiness, retail, and engineering.
Over the last year I've focused on architecting and shipping AI/LLM-powered solutions into production — agentic systems with tool-calling, RAG pipelines on vector databases, multi-provider integrations across OpenAI, Anthropic, and Gemini.
I work at the intersection of business strategy and technical execution — sitting with founders and C-levels during discovery, then leading engineering teams through delivery and long-term operations.
After shipping half a dozen RAG systems into production, I keep seeing the same retrieval failures. The model is rarely the problem — chunking and metadata are.
How I think about designing tools for agents — naming, scoping, idempotency, and why your agent keeps looping.
The reference architecture I've used to ship three production AI systems. ECS, Lambda, SQS — nothing fancy, and that's the point.
What changes when the technical decisions you make end up in a board deck three weeks later.
An agentic conversational layer embedded into a CRM. Sales directors ask questions in natural language; the agent runs SQL via tool-calling and enriches answers with RAG over historical data.
Custom CRM for a 300+ person manufacturing group with hundreds of millions in annual revenue. Centralizes commercial operations end-to-end.
Cloud platform for hospital operations — real-time tracking of medical assets and patient flow via RFID/IoT.
Technology consulting firm specializing in solution architecture and cloud-based systems.
TOTVS ERP customization and support for industrial manufacturing clients.
Brazil's leading enterprise software company.
Open to W2, C2C, or international contractor engagements. Comfortable across U.S. time zones.