Summary
Kevin Scaccia is an Artificial Intelligence Specialist with 8 years of experience building end-to-end ML systems, from modeling to production deployment. He has delivered large-scale solutions at Globo — including user deduplication across 180M+ records, GenAI-based retention email generation, and fraud-risk scoring — and has deep hands-on expertise with TensorFlow, PyTorch, HuggingFace, LangChain and custom neural architectures. Comfortable across MLOps and cloud-native workflows, Kevin implements reproducible pipelines with MLflow, Docker, GCP services (Cloud Run, Functions, BigQuery, Kubeflow) and Airflow. His background spans audio, image and sequence modeling, including one-shot voice biometrics with Siamese nets and a Triplet-Loss design he helped productionize. He pairs strong computer science foundations and advanced Python with a pragmatic product focus, often reshaping architectures to move complex research into reliable production. Based in São Paulo, he combines consulting and in-house experience to tackle high-scale data challenges that directly improve user and business outcomes.
8 years of coding experience
10 years of employment as a software developer
Master of Business Administration - MBA, Data Science & Analytics, Master of Business Administration - MBA, Data Science & Analytics at MBA USP/Esalq
Escolas Técnicas Estaduais
Bacharelado em Ciência e Tecnologia , Ciência da Computação, Bacharelado em Ciência e Tecnologia , Ciência da Computação at Unifesp - Universidade Federal de São Paulo