At Jaguar Land Rover India, I architected a reusable AWS infrastructure-as-code platform for the Vehicle Usage team. It enabled self-provisioning of AWS resources and reduced dependency on the central MLOps team by 50–60%.
I also led the design and productionization of a vehicle refueling detection system covering 70–75M journeys per month. Its scalable training and inference pipelines include feature engineering, data validation, and automated monthly retraining.
Earlier at Jaguar Land Rover India, I designed data-quality pipelines and trained an ARIMA anomaly detection model across 12B+ records for 70K vehicles. In personal projects, I built an agentic RAG financial intelligence system, fine-tuned and served an open-weight LLM, and developed a movie recommendation system.

