Joao Lucas Blum Kunzel
Data Scientist | Machine Learning EngineerI am a Data Scientist/ ML engineer with 8 years of experience transforming complex, raw datasets into high-impact, scalable AI and predictive models.
Architected and deployed an end-to-end ML production pipeline on GCP (Vertex AI) for battery preventive maintenance. Automated model runs and monitoring, projecting €M+ annual savings by reducing battery field failures. Awarded Renault's Innovation Trophy.
Engineered scalable data ingestion & ETL pipelines processing petabyte-scale vehicle telemetry. Optimized BigQuery partitioning and clustering, reducing query costs and enabling real-time model inference for fleet-wide performance.
Architected and deployed a cloud-native battery dimensioning platform on GCP, integrating real-world telemetry with physics-based simulations for precise battery sizing. Full-stack service using Python and Streamlit, utilized by Renault engineering teams to optimize battery coverage across diverse client driving profiles.
Architected the end-to-end industrialization of battery and powertrain analytics tools. Transformed raw vehicle telemetry into a scalable predictive service using Python and GCP deployment, providing automated battery SOH estimation, BMS fault estimation, and usage forecasting for millions of connected vehicles.
Developed an anomaly detection model using motor images for industrial quality control in PyTorch, enabling automated defect identification on the production line.
Let's Connect & Collaborate
If you want to geek out over Transformers, here's how we can get started:
Self-Hosted & Free Services
Privacy-focused utilities and AI experiments.