Hi, I'm Shaurya.

I'm a perception engineer. For the last seven years I've been helping robots make sense of the world: underwater, on the water's surface, in the air, and lately in the stratosphere.

It started at the DTU-AUV Lab in Delhi, where I spent my undergrad years building autonomous underwater and surface vehicles. I did a bit of everything there: mechanical design, vehicle modelling, controls, and the perception that let our vehicles find and manipulate things in murky water. Those robots took us to competition finals in India and Singapore, and one of them, VARUNA 2.0, now sits in the National Robotics Museum of India.

At WPI I did my M.S. in Robotics and went deep on state estimation and 3D vision. Two research problems from that time still shape how I think. At the PeAR Group I worked on estimating dense depth maps from reflected sound, building the acoustic raytracing simulator that generated our training data, and learned that "sensor" is a much bigger word than I thought. At the HiRO Lab I worked on making tele-manipulation easier for the human on the other end: shared camera-in-hand viewpoint control and AR cues for a humanoid tele-nursing robot, evaluated with real user studies.

Today I'm the founding engineer and computer vision lead at Space Intel AI, building StratoShield: a perception and autonomy system for geofencing critical infrastructure from drones and stratospheric balloons. I own it end to end. I picked the sensors and compute under hard SWaP constraints, built the Unreal Engine synthetic-data pipeline that let us develop perception before the hardware existed, and direct the contractors who build alongside us. Architecture in the morning, vendor calls in the afternoon, debugging coordinate frames at night.

If any of it resonates — or you just want to argue about quaternion conventions — my inbox is open: shauryaparashar@gmail.com.