- Developed a scalable Next.js/Golang Application to provide a centralized dashboard for monitoring and managing multiple GPU/data resources, offering granular role-based-access controls enhancing security

- Worked collaboratively with the Unity team, ensuring seamless integration of features and tasks, while contributing to the development of Large Language Models (LLMs) to enhance AI capabilities within the project and use-cases.

- Conducted in-depth research on Large Language Model (LLM) architectures. Implemented RAG pipelines to enhance LLM performance and accuracy, Leveraging innovative techniques to achieve more precise and faster responses

- Implemented a high-performance gRPC framework to facilitate communication between Unity and a Python back-end, utilizing Jupyter and Docker, enabling the exchange of complex data structures and real-time updates.

- Implemented Behavior Trees Approach to automate the finite machine’s instructive tasks having access to LLMs.

- Integrated ChatGpt with prompt-making React App offering speech recognition and translating from multiple languages, Converting given instruction into robotic actions for planning the execution on real-world robots.

- Comparative study done on different path Planning Algorithms suitable for Autonomous Vehicle Driving in Indoor and Outdoor Environments, Coded (in C++ and python) and Optimised for same.

- Trained Custom Segmentation model on Recorded Video footage to detect footpaths for Last Mile Delivery Robots.

- Designed end-to-end (Warehouse To Delivery Location) path planning algorithm in C++ to deliver in min time.

- Implemented Multi-Robot Cloud Architecture on Kubernetes to test Cluster network communication for robots.

- Created An OSM Planner which accepts both gui and terminal inputs, generating path over pixelated Open Street Map tiles with calculated approx. distance from Indoor location to Outdoor Goal and Vice-Versa Considering switching between different modes of driving. Task Achieved in C++, using Xml parser lib, Astar algorithm, OSM data, Rviz.

- Recreated 3d World of targeted Area in real world for Autonomous Driving using OSM Model and Blender plugins.

- Implemented Cloud Robotics Architecture on Kubernetes. Successfully tested Cluster network communication through DDS layer of Ros2 Framework. Deployment Achieved in both minikube and microk8s

- Worked on Ros based robots local navigation and configuration parameter tuning for custom made robots.

- Robotic 5 Dof Arm Design in Fusion 360 and 3d Printed same for prototype testing with actuators and Moveit Motion planning Framework of ros system.

- Explored Research domains under ISRO Lab in Metallurgy and Nano-Technology for advanced materials for aerospace applications.

- Participated in Robotics and AI competitions and hackathons, gaining hands-on experience in automation, machine learning, and collaborative problem-solving.