Anadrones System Pvt. Ltd.

• Designed swarm intelligence algorithms for military-grade QinetiQ Banshee target drones; coordinating 5+ agents with real-time trajectory and motion planning at speeds exceeding 200 km/h.
• Built ROS2 control architecture with fault tolerance and dynamic reconfiguration under DO-178C safety standard; reduced critical software defect rate by 35%.
• Developed Digital Twins of UAV platforms in NVIDIA Isaac Sim and Omniverse with high-fidelity point cloud and sensor modeling, reducing field-testing iterations by 40%.
• Containerised simulation pipelines using Docker; integrated CI/CD workflows cutting deployment time by 50%.
• Tools: ROS2, NVIDIA Isaac Sim, Omniverse, Docker, C++, Python, DO-178C

People Tech Group Pvt. Ltd.

• Led development of control, perception, and motion planning stack for an autonomous UGV using ROS2 + SLAM; achieved sub-10 cm localisation accuracy in unstructured environments.
• Developed real-time C++ (17) modules for UGV motion control at 100 Hz loop frequency integrated with low-level sensor drivers via ROS2 nodelets.
• Improved UGV localisation accuracy by 30% using computer vision on Google Street View features; processed point cloud data from VLP-16 LiDAR at 10 Hz.
• Implemented sensor fusion (LiDAR, RGB-D, IMU) with CUDA-accelerated preprocessing; reduced SLAM CPU load by 25% while maintaining real-time performance.
• Managed a team of 8 engineers across 6 milestone deliveries; maintained 100% on-time delivery record using Agile/Scrum.
• Tools: ROS2, C++17, Python, CUDA, Gazebo, LiDAR, IMU

Indian Institute of Technology

• Developed and deployed distributed adaptive coverage control algorithms on a 6-UAV testbed using ROS + OptiTrack Motion Capture; achieved <5 cm positioning error indoors.
• Led UGV mapping project integrating Extended Kalman Filter + deep learning for semantic terrain understanding; mapped 3 terrain types at >95% segmentation accuracy.
• Scaled multi-robot simulation environments using RTOS-compatible control loops; coordinated between faculty, PhD students, and industry stakeholders.
• Sample Videos
• Tools: ROS1/ROS2, Python, EKF, Deep Learning, OptiTrack, RTOS

Veda Aeronautics

• Designed Visual Inertial Odometry (VIO) navigation stack in C++ with tightly-coupled Kalman filters on Jetson Nano; achieved <2% drift over 100 m trajectory in GPS-denied conditions.
• Built reference-based navigation system enabling a micro drone to track targets at up to 15 m/s in both GPS and GPS-denied environments using ArduPilot/Pixhawk flight stack.
• Collaborated with hardware and software teams to integrate the system into a fully functional autonomous robotics solution.
• Tools: C++, ROS, VIO, Kalman Filters, Jetson Nano, ArduPilot, Pixhawk, OpenCV

IISER Bhopal

• Developed Model Predictive Control (MPC)-based cooperative path following for autonomous surface vehicles (ASVs) in a multi-agent ROS framework.
• Built OFFSEG framework for off-road semantic segmentation on RELLIS-3D and RUGD datasets; improved scene understanding for off-road vehicle navigation and decision-making.
• Led underwater SLAM system configuration and performance tuning on BlueROV2 under real water-testing constraints.
• Tools: ROS, Python, CasADi, MPC, Deep Learning, BlueROV2

AIOTIZE

• Implemented ROS-based drone frameworks for periphery monitoring and asset tracking.
• Built a Classification and Detection Model using TensorFlow; optimised and deployed it on Nvidia Jetson TX2.
• Developed a Data Relaying System using drones as a communication channel over ROS.
• Tools: ROS, Python, TensorFlow, Jetson TX2, OpenCV

Languages and Databases

Python

C++

C

Embedded C

Java

Shell Scripting

Libraries

NumPy

Pandas

OpenCV

scikit-learn

matplotlib

Frameworks & Middleware

Keras

TensorFlow

PyTorch

ROS 1

ROS 2

DevOps & Other Tools

Git

AWS

Docker

CUDA

RTOS

DO-178C