Yago Costa Ayala
Open to US remoteAI Software Engineer · Santa Catarina, Brazil (UTC-3)
I build production LLM features — RAG, agents and conversation memory — in TypeScript and Node.js. Before that, four years of backend and full-stack.
Selected results
- 30K+users
- served by the RAG, agents and memory layer I built
- Clinicorp · 2025–26
- 600K+integrations
- had failed silently. I traced them to one root cause and brought them back
- LIVE Roupas Esportivas · 2023
- 1000+clients
- migrated to a new system with zero downtime. Nobody noticed
- Clinicorp · 2022
Core stack
- TypeScript
- Node.js
- OpenAI API
- RAG
- Python
- Go
- React
- Next.js
- PostgreSQL
- AWS
- GCP
About
AI Software Engineer with 5+ years of experience. I build production LLM features in TypeScript and Node.js — RAG, agents and conversation memory — for a product used by 30,000+ people. Today I work on the broker-facing web app of a US freight brokerage platform at DAT Freight & Analytics.
My first four years were backend and full-stack, where I traced 600,000+ failed integrations to a single root cause and ran a zero-downtime migration for 1,000+ enterprise clients. I work from Brazil on UTC-3 and cover the full US East Coast day. English C2 (EF SET 78/100).
Stack
- AI / LLM
- OpenAI APIHugging FaceRAGAgentic WorkflowsConversation MemoryLLM EvalsGuardrailsPrompt EngineeringCost/Latency Optimization
- Languages
- TypeScriptJavaScriptPythonGoPHPSQL
- Backend
- Node.jsExpressREST APIsGraphQLMicroservicesWebSockets
- Frontend
- ReactNext.jsTailwind CSSRedux
- Cloud / DevOps
- AWSGCPDockerCI/CD
- Databases
- PostgreSQLMySQLMongoDBRedis