At Banco Votorantim, I developed autonomous agents that generate test scripts from natural-language scenarios, reducing manual effort in software validation. I integrated large language models and led the work from agent architecture through implementation and fine-tuning.
At Hapvida, I architected and built a Google Cloud data pipeline with a medallion architecture, from raw data processing to business-ready dashboards. I owned the project through production deployment.
At Capco, I’m leading the data ingestion layer of a cloud migration to Microsoft Azure, building pipelines with Databricks and Azure Data Factory. Earlier, at Mercedes-Benz, I contributed to a project that reduced production-system errors by 20%.
