Apr 2024 – Jul 2025 · 1 yr 3 mos · Bengaluru, Karnataka, India · On-site

• Lead CV/ML Architect of
• Perception features for ADAS in OLA Electric Scooters.
• AI Solutions for Real-Time Video Analytics.
• Perception & Data Management Stack for building OLA Maps.
• Pursued active research on Multi-task End-to-end Perception Networks & Foundational Models for Autonomous Driving, aiming to advance scalable, real-time AI for next-generation mobility in India.
• Additionally, developed, trained and deployed
• Lane Detection and Departure Warning System for ADAS in OLA Electric Scooters.
• An End-to-end Detection & OCR Model for Automatic Number Plate Recognition (ANPR) to streamline online vehicle registration.
• Instance Segmentation Models for UR10 Robotic Arm Pick-and-place in Stores & Warehouses to automate online order processing.
• Semantic Segmentation Models for drive scenes in Unstructured Traffic and Adverse Weather Conditions.

Mar 2024 – Apr 2024 · 1 mo · Bengaluru, Karnataka, India

• Started R&D on Perception for ADAS and Autonomous Driving.

Nov 2021 – Sep 2023 · 1 yr 10 mos · San Diego, California, United States

• Led Systems, Simulator and API development for the 3D Reconstruction feature set for the next generation of Snapdragon XR2+ Chipset.
• Developed 3D Reconstruction algorithms on the Snapdragon XR platform cDSP, leveraging Qualcomm Hexagon ISA and HVX.
• Implemented LE AArch64 optimizations for a 6DoF Tracker, improving power efficiency and computational performance.

Feb 2020 – Nov 2021 · 1 yr 9 mos · San Diego, California, United States

• Led Systems, Simulator and API development for key Computer Vision features, including Optical Flow, Depth from Stereo, Feature Point Extraction, Image Pyramids & Downscaling, on Qualcomm’s Engine for Visual Analytics (EVA 2.1/3.0) - a Hardware Accelerator for Computer Vision algorithms on Snapdragon XR Gen 1 & 2 Mobile platforms.
• Designed and deployed a Continuous Integration and Testing System using Jenkins (later on TeamCity) for automated testing and version/patch releases of the Simulator across Windows, Linux and Android platforms, supporting multiple x86-64 and AArch64 build variants.

May 2019 – Aug 2019 · 3 mos · San Diego, California, United States

• Developed Systems, Simulator and API for Qualcomm’s Computer Vision Processor (CVP 2.0) - a Hardware Accelerator for Computer Vision algorithms on Snapdragon Mobile platforms.
• Enabled SDK verification and testing using Jenkins.

Aug 2019 – Dec 2019 · 4 mos · Pittsburgh, Pennsylvania, United States

Jan 2019 – May 2019 · 4 mos · Pittsburgh, Pennsylvania

Dec 2018 – May 2019 · 5 mos · Pittsburgh, Pennsylvania

• Served as a Joint Funding Committee (JFC) Financial Advisor, supporting the allocation of $2 Million in student funds across 300+ clubs and organizations.
• Advised an assigned subset of clubs and organizations, reviewing and refining their annual budget proposals.

Jan 2019 – Dec 2019 · 11 mos · Pittsburgh, Pennsylvania, United States

• Advisor - Dr. David Held
• Developed 3D Semantic Segmentation models for Automatic Powerline Inspection using LiDAR point cloud data.
• Developed Object Detection and Semantic Segmentation models for Aircraft Defect Detection using RGB camera data.

May 2018 – Jul 2018 · 2 mos · Bangalore Urban, Karnataka, India

• Developed and implemented a robust 6-DoF Positional Tracking algorithm for XR (Extended Reality) applications, enhancing performance under motion blur through Spatially-varying Blur Kernel Estimation.

May 2017 – Aug 2017 · 3 mos · Hong Kong SAR

• Advisor - Dr. Pan Hui
• Laboratory - HKUST-DT System and Media Laboratory (SyMLab)
• Developed a Real-Time Visual Privacy Protection pipeline for pedestrians, integrating a customized YOLO-based Network for detection, Correlation Trackers for tracking, and DeepID CNN features for recognition.

Jan 2017 – May 2018 · 1 yr 4 mos

• Advisor - Dr. K Madhava Krishna
• Laboratory - Robotics Research Center (RRC)
• Developed a novel 3D Keypoint Localization Network based on a Stacked-Hourglass Pose Estimation Network, leveraging shape priors to model wireframes around cars in dynamic road scenes using monocular camera input.
• Designed and implemented a real-time Road Sign Detection & Classification system using modified LeNet and AlexNet CNNs, contributing to the Autonomous Driving Stack as part of the university’s team for the Mahindra: Spark the Rise Driverless Car Challenge.

Oct 2015 – May 2018 · 2 yrs 7 mos

• Advisor - Dr. C V Jawahar
• Laboratory - Centre for Visual Information Technology (CVIT)
• Developed an AI-driven feedback system for students, by encoding key scholastic performance features, followed by an Auto-regressive LSTM Decoder.
• Designed and built an Android application for scanning and processing handwritten assignments, enabling Automated Plagiarism Detection and Neatness Evaluation through ML models running on a Django server.
• Developed a novel Bag-of-Words architecture for Handwriting Aesthetic Analysis, utilizing a modified SIFT+SURF feature extractor, GMM Distribution, and Fisher Vector Encoding, followed by an SVM Regressor.

Sep 2015 – Nov 2015 · 2 mos · Hyderabad, Telangana, India

• Developed a lightweight background application using Java and Sentry, designed to minimize resource consumption while analyzing the performance of running applications and processes through real-time diagnostics and crash report generation.

May 2015 – Jul 2015 · 2 mos · Princeton, New Jersey, United States

• Developed custom iOS 8 applications using Objective-C, tailored to client-specific requirements. Notable projects include a flexible contacts and user information manager with custom export capabilities, among others.