Senior backend engineer, 13+ years across SaaS, fintech, and government tech. Brazilian Portuguese name pronounced ee-NAH-see-oh. Based in Recife, working remote with US/EU teams.
I spend most of my day on production AI systems: multi-provider LLM orchestration, agentic workflows on LangGraph, RAG pipelines, and the eval discipline that keeps them safe to ship in regulated domains.
I also operate an autonomous-agent control plane on a SaaS I built solo. Claude Code agents shipping multi-step migrations while I sleep. Most recent unsupervised run was a three-phase PII encryption migration with key rotation and backfill. It worked.
- Languages: Python, Java, TypeScript, JavaScript, SQL.
- Backend: Django, DRF, Celery, Spring Boot, Spring Data JDBC, FastAPI, Node, Express.
- Frontend: SvelteKit, Next.js, React, Capacitor.
- Data: PostgreSQL, PostGIS, MySQL, Redis, Qdrant, Elasticsearch, Cassandra, BigQuery.
- AI / LLM: LangChain, LangGraph, OpenAI, Anthropic, Google Gemini, Perplexity, Voyage AI rerank, PyMuPDF, GPT vision, MCP servers.
- Infra: AWS (Lambda, EventBridge, Step Functions, SNS+SQS, Aurora, ECS, S3), GCP (Dataflow, BigQuery, Pub/Sub), Kubernetes, Docker, Terraform.
- Designing golden-set eval regressions for regulated-domain LLM systems.
- Multi-provider LLM circuit breakers (using structured-output validation rate as a health signal).
- Building MCP servers that proxy into an existing REST API.
- Surviving 6.5 years at Liferay across five roles.
