Marcos Paulo Freitas

At Google, I built a Python internal framework that identified, filtered, and grouped more than 50 performance optimization opportunities across large-scale C++ systems.

I curated optimization cases for few-shot LLM-assisted code analysis and refactoring, and integrated CPU and memory profiling into automated diagnostic pipelines. My work investigated runtime, heap fragmentation, algorithmic complexity, lookup, and serialization bottlenecks before delivering an experimental tool for continued engineering use.

Previously at Dito CRM, I architected and maintained large-scale data structures and cloud-based warehouses using SQL and data-processing frameworks for mission-critical services.

I also optimized backend logic, supported real-time event tracking, resolved architectural bugs and performance bottlenecks, and collaborated across the full software development lifecycle.