Ing. Krishnajith Gokul

Built a synthetic data generation pipeline bridging Valeo’s in-house SIL toolchain with Unity, generating large sequences of diverse vehicle motion use cases by defining scenarios, triggered in a single command Cut initial data collection time drastically: campaigns that typically take 2 to 3 Working days for 100 to 200 traces now take 4 to 5 hours using the simulator Tool Development: Established Git version control for the ML codebase, built a GUI for the pipeline to visualize the trained model’s predic

Automated the full workflow from project initialization through training to deployment on GCP Vertex AI, triggered by a single command and containerized with Docker for reproducible runs Designed and implemented a full ML pipeline in Python and PyTorch covering data strategy, labelling methodology, preprocessing, training and evaluation Researched and deployed Deep Learning architectures (Set Transformers, RNN/GRU encoder decoder, point cloud decoders) for ultrasonic feature extraction, training on GCP Ve

Semantic Segmentation of Synthetic Aperture Radar (SAR) Images in Autonomous Driving Scenarios

Proactive testing and simulation of ADAS systems, including Automatic Parking Assist and Navigation Pilot etc. Identified and meticulously documented system anomalies for ongoing refinement of the functions Utilized advanced visualization tools for detailed scenario classification. Developed and executed complex test cases to stress-test the planning capabilities of autonomous systems in accordance with Euro NCAP guidelines. Prepared vehicles for customer demonstrations, ensuring optimal performance an

Works in Supply Chain and Logistics Management Domain providing consultancy to e-commerce division of H&M,primarily involving inventory management , inbound outbound shipping processes and day to day interaction with Third party logistics across the globe.