Matheus Marini

Computer Science student at Centro Universitario Senac (graduating December 2026), focused on Python backends and applied AI. I was the top contributor to an LLM recommendation agent built with LangGraph: tool-calling agent, model benchmark, human-in-the-loop approval, Pydantic-validated output and a 90-call prompt A/B test showing that prompt-injection defense must live in code. For my thesis I trained YOLOv8 detectors in PyTorch with 5-fold cross-validation. I also built an online code judge (FastAPI, PostgreSQL, Redis queue, workers running code in isolated Docker containers) and a ticketing platform with 273 backend tests, using generative AI in my workflow while reviewing and testing the output. At Apple Developer Academy | Senac I spent nearly two years on an agile team with code review and production troubleshooting, shipping 4 of our 7 apps to the App Store. I am also taking the AWS Cloud course at Escola da Nuvem and built a CI/CD pipeline on AWS. Advanced English, native Portuguese.