Hi!

I’m a technical Research Practitioner working in Deep Learning and Perception-focused areas. My current interests include deep learning, generative modeling, perception and motion planning, with an emphasis on building robust and deployable solutions. I’ve extensive industry experience, working closely with leading automotive clients. During this time, I’ve also gained research experience, while working on multiple patents and publications, leveling myself up for the production grade solutions. Previously, I’ve graduated in Robotics from University of Michigan, Ann Arbor. I’ve worked towards:

• Implementing a Large Language Model (LLM) from scratch using PyTorch, including custom tokenization, embedding layers, multi-head self-attention, transformer blocks and autoregressive training pipelines to analyze scalability, convergence behavior and inference dynamics: Large Language Model from scratch
• Studying the Diffusion based generative models setup on the benchmark Transformer architectures (Transfuser++, Interfuser) and Vision-language models (LMDrive, CarLLaVA) to utilize them for generating adversarial scenarios to predict violations
• Development for Perception Systems including Object data clustering, Classification and Target tracking under Lidar, Radar and Event camera system with Deterministic and ML components
• Deployment of safe controllers for building safety critical features like Adaptive Cruise Control (ACC) and Autonomous Emergency Braking (AEB) systems

You can also find my work on my google scholar page. In parallel, I’ve been building a beginner-friendly research repository for Deep Learning algorithms with minimal dependencies, to create foundations for a more evolved project.

I also enjoy meeting new people, share perspectives and engage into collective brainstorming sessions. Hit me up if you have any project idea, want to collaborate or just chat!