Working with AWS customers for implementing Gen AI solutions

- Developed co-pilot agents leveraging Large Language Models to enhance Ridecell's data management capabilities- Integrated custom Retrieval Augmentated Generation (RAG) pipelines to enrich data augmentation process with LLMS- Fine tuned LLMs with Parameter efficient fine tuning (PEFT) like LoRA for better task specific performance- Evaluated RAG pipelines and LLMs to improve response and reduce hallucinations.- Designed and deployed scalable, reliable, and available Deep Learning models- Lead MLOps Initiatives to streamline ML training and deployment pipelines- Analyzed thousands of hours of data from automotive fleets and vehicles, extracting vital information, metrics, and features using statistical and machine learning techniques.- Leveraged sensor data for scenario extraction and metrics evaluation, contributing to the development of advanced analysis tools like nemosearch.ai. (https://nemosearch.ai/)

- Research, Development, Implementation and Testing of Motion Planning Algorithms for autonomous driving- Working together with Perception, Prediction, Mapping, Localization and Control systems teams- Working with LIDAR, Cameras, GPS, IMU and other sensors for autonomous driving- Autonomous driving full stack testing, debugging and maintenance

Navitgation, motion planning, software development, Deep Learning, Prediction

Course - "Mechanics of Locomotion"

Space and Legged Robotics Labratory

Analysis and implementation of human walking gait on bipedal systems.