Vikhyat Chauhan
vikhyatchauhan@vt.eduContact Me
About
I’m Vikhyat—a builder at heart who loves turning ideas into reliable products. My recent work: an NSF-backed UAV stack that cut deadline misses by 80% and reduced safety violations by 23%—without slowing down. Earlier, at GE Healthcare, I helped modernize real-time imaging platforms. I care about clean, well-documented code, meaningful metrics, and tools that help teams move faster. When I’m not coding, I’m probably meal-prepping a new vegan recipe or sketching the next idea. If you’re building at the edge of autonomy and AI, let’s talk.
Experience
Virginia Polytechnic Institute and State University
Virginia, USA
AI/ML Software Engineer
Nov 2024 - Present
- Delivered an NSF Funded UAV control stack combining learning-based perception with a deadline-aware planner (mixture-of-experts); reduced deadline misses by 80%, lowered zone violations by 23%, and kept median time within 2% of the fastest policy—validated via simulation and statistical analysis.
- Developed a SAC-based controller with continuous control and reward shaping, leveraging GPU training, domain randomization, and ROS2 middleware; productionized with Docker, CI, and metrics/ablation tooling to validate robustness in constrained environments.
GE Healthcare
Bangalore, IN
Full Stack Software Engineer
Mar 2022 - Jun 2024
- Delivered 20% faster MRI segmentation on a GE real-time MRI pipeline by integrating Medical Stitching with ViT/Swin transformers and end-to-end TensorRT optimization for clinical use.
- Recognized with the GE Impact Award for reduced MRI deployment time by 80% by migrating a C++ monolith to Kubernetes microservices, enabling scalable, zero-downtime rollouts to PACS/RIS; shipped MRPasting and GE Edison–optimized models via Docker + CI/CD (Jenkins, Git, SonarCloud).
TNM Electronics
Gurugram, IN
Technical Product Manager
Sep 2021 - Mar 2022
- Drove end-to-end delivery as PM/tech lead at a seed-stage startup—coordinated hardware, firmware, and cloud teams (plus vendors), aligned priorities, and unblocked dependencies to launch the AWS IoT platform on schedule (100+ devices, 1K+ events/day, 99.9% uptime)—highlighting a collaborative, customer-first approach.
- Built a resilient IoT platform on AWS EC2 with 99.9% uptime, scaling to 100+ devices / 1K+ events per day via RabbitMQ, MongoDB, and a custom MQTT broker integrated with patented hardware.
Mahindra Susten
Maharashtra, IN
Research Intern
Jun 2018 - Mar 2020
- Built a Linux-based autonomous quadcopter controller (MAVLink + advanced control) and integrated it into Mahindra Susten’s solar ops—automating routine inspection flights and cutting manual intervention across sites.
- Shipped an ML-driven thermal fault monitoring + end-to-end drone inspection pipeline that automated detection and reporting, delivering a 35% reduction in labor costs and faster issue triage.
Universal Robots
Maharashtra, IN
Research Intern
Jun 2018 - Mar 2020
- Deployed Universal Robots cobots across multiple manufacturing lines—designed safe workflows, trained operators, and optimized cell layouts—boosting throughput and reducing changeover time while meeting safety standards.
- Built production-grade automation with URCaps and custom scripts (pick-and-place, machine tending, QA), creating maintainable routines that cut manual variability and unplanned downtime.