At ROIT Solutions, I model and maintain BigQuery analytical bases for tax data, and optimize queries and pipelines. One main flow saw processing time fall by 45% and costs by 32%.
I built an AI-powered tax classification platform that combines LLMs, business rules, and semantic caching. It uses FastAPI services and BigQuery persistence, with auditability and AI cost monitoring.
At ROIT Solutions, I also developed Python pipelines and automations integrating APIs and cloud tools, and contributed to tax analyses that identified approximately R$ 1.2 million in opportunities.
Previously, as an Engineering Assistant — Data, BI & Continuous Improvement at TechnipFMC, I developed a supplier evaluation solution that reduced inspected items by 40%–60% and inspection hours by 30%–50%. I also digitized processes with Power Platform tools and developed operational and quality dashboards in Power BI.
