At InCreative Labs, I architect and deliver production AI platforms across backend, web, mobile, and infrastructure. I built a multi-provider LLM orchestration layer with semantic routing to select models based on intent and complexity, reducing cost per query.
I developed RAG and hybrid search pipelines with PostgreSQL and pgvector, combining full-text and vector search through Reciprocal Rank Fusion. I also implemented LLM cost observability with Langfuse, price versioning, and content-hash caching.
Previously, at NTT DATA Europe & Latam, I led the design and implementation of multi-tenant APIs and payment services for BEES. Across my career, I’ve delivered software for fintechs, government agencies, and global technology companies, and co-founded Metrópole Interativa, where I led technical operations.
