ML Engineer · MLOps & Cloud AI (AWS) · Brazil · Remote
I build and operate ML/NLP pipelines and event-driven serverless architectures in production on AWS.
Currently open to ML Engineer / MLOps roles in structured teams.
At Verzel (AI Developer — Lunella/RosanaDesk SaaS):
- Architected event-driven serverless system on AWS: API Gateway → SNS → SQS FIFO → Lambda → EventBridge — 9+ microservices in production
- Built Python ML/NLP pipelines on AWS Lambda for topic extraction and classification from conversational data, with multi-tenant consolidation and automated PDF/Slack reports
- Integrated LLM agents (TaskingAI) and voice assistant pipeline (VAPI — STT→LLM→TTS)
- Reactive Java backend: Quarkus · PostgreSQL · Redis · Keycloak · Flyway
Research (two master's degrees in progress):
- UFABC — adaptive control (LQR, H∞), fault detection (EKF/UKF/CUSUM), LLM agent supervisor (CrewAI) in Hardware-in-the-Loop
- UFPA — model-free predictive control and Deep RL/DQN for power electronics
|
NLP pipelines · topic extraction · model serving · drift monitoring · MLflow |
Docker · CI/CD · Pytest · versioned migrations · reproducible deploys · model monitoring |
Lambda · SQS · SNS · EventBridge · S3 · PostgreSQL · Redis · pgvector |
RAG · LangGraph · CrewAI · prompt engineering · Langfuse · STT→LLM→TTS |

