Bengaluru, Karnataka, India
I build the software infrastructure that enables autonomous driving systems to be validated at scale. As an SDE-II at Jaguar Land Rover, I work on the nervous system behind ADAS - middleware, signal orchestration, and high-performance data pipelines where reliability and low latency are critical. My primary tools are C++ and Python, with a strong focus on systems engineering and scalable architecture. Some of the systems and infrastructure I’ve been working on: • Engineering a C++ SOA-to-MF4 conversion tool for vehicle network data, including SOME/IP parsing, transport protocol fragmentation/reassembly, multi-message packet handling, and dynamic payload decoding. • Architecting high-throughput pipelines that convert large-scale simulation and logging data into MF4 measurement formats for downstream validation and analysis. • Building signal-processing frameworks that reduced redundant computation by 80%, accelerating perception-feature deployment and validation cycles. • Designing centralized signal-definition infrastructure to improve consistency and interoperability across ADAS workflows and engineering teams. Previously, I worked on automation and simulation tooling for ADAS validation, including Unreal Engine-based models and automated AEB evaluation workflows using Simulink and CarMaker. I also spent time building real-time computer vision systems at Samsung R&D. I'm an IIT Bombay alumnus with a background in Electrical Engineering and Mathematics. I enjoy working on modern C++, middleware architecture, distributed systems, and the invisible infrastructure behind autonomous systems.
ADAS Infrastructure & Middleware • Built a high-performance signal-processing framework in C++ and Python that reduced redundant computation by 80%, accelerating perception-feature deployment and validation workflows. • Engineered a C++ SOA-to-MF4 conversion and diagnostic tool for vehicle network data, including SOME/IP parsing, transport protocol fragmentation/reassembly, multi-message packet handling, and dynamic payload decoding. • Architected scalable serialization pipelines for converting large-scale simulation and logging outputs into MF4 measurement formats for downstream validation and analytics. • Designed a centralized Signals Database to standardize ADAS definitions and improve interoperability across engineering workflows and global teams. • Developed scalable Python parsers for OpenDRIVE maps, enabling automated generation of structured ground-truth data for large-scale simulation environments. Earlier Work in Validation & Simulation • Automated AEB testing in Simulink and CarMaker, reducing evaluation time by 90%. • Developed interactive Unreal Engine models for Park Assist, doubling validation speed for HMI workflows. • Mentored new engineers on simulation workflows and refactored tooling for the wider systems engineering team.
Developed a low-light image enhancement and deblurring algorithm using residual deconvolution, successfully improving restoration for blur kernels up to 17×17. Built an efficient overexposure correction model to refine light source glow, achieving real-time processing (<1s) for 512×512 resolution images.