I build technical systems that hold up when you look closely.

B.Tech Mathematics & Computing student at Central University of Jammu. Focused on real-time computer vision, edge machine learning, and competitive robotics: measured, benchmarked, and built to be inspected.

1st Place RoboTraverse '261st Place Le Mansar '26
PythonC / C++OpenCVMediaPipeESP32 / FreeRTOSPyTorchLinux Kernel (uinput)MQTT / IoTNumPyComputer VisionDifferential DriveAstro / TypeScriptPythonC / C++OpenCVMediaPipeESP32 / FreeRTOSPyTorchLinux Kernel (uinput)MQTT / IoTNumPyComputer VisionDifferential DriveAstro / TypeScript
Focus

Real-Time Computer Vision · Edge AI · Low-Latency Robotics · Embedded Firmware

Education

B.Tech in Mathematics & Computing · Central University of Jammu (2025–2029)

Current Status

Mentoring student robotics teams at CUJ, e-Yantra competition build work, & edge CV research

Featured Builds

Selected Work

Full Project Archive →
1stRoboTraverse Club Champions
~40 FPSCPU-only hand tracking throughput
<25 msEnd-to-end Wayland input latency
Engineering Philosophy

How I work

01 // Frame

Define constraints before writing code

Quantify latency bounds, CPU utilization ceilings, memory footprints, and explicit success criteria before choosing a stack or architecting a pipeline.

02 // Build

Make it measurable from day one

Construct the smallest functional end-to-end slice that can be profiled. Hardware in the loop, telemetry streamed early, assumptions verified on real silicon.

03 // Profile & Refine

Isolate and eliminate real bottlenecks

Profile frame jitter, kernel context switches, and power draw. Document the actual engineering tradeoffs made between latency, responsiveness, and thermal stability.

Background

Experience & Mentorship

All Experience →

Tech Mentor

Central University of Jammu · e-Yantra

  • Mentoring student technical teams across robotics, embedded systems, algorithms, and software challenges.
  • Guiding hands-on build work around ESP32, wireless telemetry, and real-time sensor processing.

Summer Intern · AI & ML

ShadowFox

  • Built and evaluated a custom CNN workflow on CIFAR-10 using deep learning tooling and GPU acceleration.
  • Analyzed 8,399 retail orders with Python, Pandas, and Plotly to study segment-level margin and absolute profit.
  • Explored Gemma 2 attention behavior and compared observations with GPT-2.
Open for Opportunities

Have something worth building?

Open to internships, research roles, and technical collaborations in computer vision, robotics, and embedded systems where shipping real code is part of the work.