Jun 2020 – Jan 2022 · 1 yr 7 mos

• Contributed to the research and development of below projects.
• 1. ISAAC SIM, DRIVE SIM - NVIDIA's simulation platform based on Omniverse for robotic systems and autonomous vehicles.
• 2. My primary research focus is around Sim2Real transfer learning to address the domain gap problem via techniques like domain randomization, self-supervised learning, domain adaptation.
• My area of work is at the intersection of robotics, computer vision, simulation and deep learning.

May 2018 – May 2020 · 2 yrs

• Project Isaac Sim: Robotics, Simulation and Deep Learning

Aug 2017 – May 2018 · 9 mos · Los Angeles, California

• Advised by Professor Gaurav Suhkatme and Professor Lantao Liu
• Designing and developing novel reinforcement learning algorithms for efficient path planning for underwater marine vehicles.

May 2017 – Aug 2017 · 3 mos · Thousand Oaks, California

• 1. Automated Vision Inspection via Deep Learning: I was the technical lead for the project. I developed a machine learning algorithm using Google’s Tensorflow that can adaptively learn to classify between defective/ non-defective images of syringe or vials. This approach enhanced the vision system by reducing the number of ejects by around 30% which eventually saved both time and money for Amgen.
• 2. Virtual Tour: Developed a generic framework for virtual tour using Unity that can be deployed in multiple platforms like iOS, Android mobile app or deployed as a WebGL app. This can be leveraged for orientation for new hires or general demonstration to any audience.
• 3. 3D Molecule Visualization: Contributed to a tool named UnityMol using Unity and C# for scientists at Amgen to visualize and interact with molecule in a virtual environment using VR devices like HTC VIVE, Oculus Rift.

Sep 2016 – May 2017 · 8 mos · Los Angeles, California

• Advised by Professor Maja Mataric
• 1. Contributed to the development Kinect Bridge that enables transfer of messages from Microsoft Kinect on the Windows end (Server) to ROS messages at the Ubuntu end (Client).
• 2. Build computational techniques that will enable social robots to encourage social and cognitive growth in children, including those with social or cognitive deficits.

Jul 2014 – Jul 2016 · 2 yrs · Hyderabad Area, India

• 1. Cloud Based ETL Framework: Developed a cloud based solution on Azure that reduced the deployment time for ETL related entities by 80%. Gave a talk at Microsoft’s Machine Learning, Analytics and Data Science (MLADS) Conference.
• 2. Let's Envisage: Lead a team of 4 to build a tool that smartly assigns work items to members of team for particular sprint using machine learning. The project was started by me as a side project in order to solve a day-to-day problem in software industry that eventually drove interest among the Microsoft community globally. I eventually filed a patent for the same at Microsoft.
• 3. Manage Rewards and Promotions Tool: Contributed to Microsoft's appraisal tool which is currently used by 20000 managers across Microsoft. I contributed to backend application development using SQL Server 2014 and service layer using .NET and Visual C#.

Jan 2014 – Jun 2014 · 5 mos · Munich Area, Germany

Jun 2013 – Aug 2013 · 2 mos · Hyderabad Area, India