About me

Yash Bansod A highly motivated, efficient and organized individual with an inquisitive mind and a passion for learning new skills. Seeking to leverage my out of the box thinking, and problem-solving skills to grow in the role of artificial intelligence engineer. My experience with robotics, artificial neural networks, operations research, and embedded systems can be efficiently used in the development of novel artificial intelligence solutions.

Download my CV


University of Maryland

College Park, MD, US
Master of Science in Systems Engineering, Robotics Specialization
August 2019 - May 2021
GPA – 4.0/4.0

Manipal Institute of Technology

Manipal, KA, India
Bachelor of Technology in Mechatronics Engineering, Robotics Specialization
July 2014 - July 2018

Work Experience

Naval Research Lab Research Group, University of Maryland

College Park, MD, US

Graduate Research Assistant (Artificial Intelligence)
Jan 2021 – Present
  • Researched an iterative graph traversal based re-entrant Hierarchical Task Network Planner. (AI planning, python)
  • Invented an integrated planning and acting algorithm that provides ~20% improvement in planning and ~30% improvement in acting performance. (Integrated AI planning and acting, graph theory)

Tubaldi Lab – University of Maryland

College Park, MD, US

Graduate Research Assistant (Machine Learning)
April 2020 – Jan 2021
  • Enhanced the speed of finite element analysis dataset generation by ~800% using concepts of distributed computing and process parallelism. (Shell scripting, multi-processing, python).
  • Established orders of magnitude speedup in the inverse structural design of meta-materials by architecting an optimization algorithm using generative neural networks. (Deep learning, python, TensorFlow)
  • Proved efficacy of neural networks in prediction of finite element calculations by replicating state of the art deep learning algorithms for forward prediction of structural properties. (Deep learning, python, tensorflow)
  • Improved the learning time of generative inverse design networks by ~300% using active learning strategy.

Continental Automotive

Bengaluru, KA, India

Machine Learning Software Engineer
Aug 2018 – July 2019
Machine Learning Intern
Jan 2018 – June 2018
Computer Vision Intern
May 2017 – July 2017
  • Improved the average training speed of convolutional neural networks in Continental's tensorflow based deep learning framework by ~100% by:
    • Implementing data pipelines enabling optimized ingestion and pre-processing of huge datasets.
    • Implementing data pipelines enabling optimized ingestion and pre-processing of huge datasets.
  • Demonstrated data-oriented behavior, path, and motion planning by conceptualizing a recurrent neural network-based planner architecture. (Deep learning, planning, python, TensorFlow).
  • Established a baseline to compare the neural network planner against conventional path planning algorithms like A*, ARA*, D*lite, and RRT. (Graph theory, python, C++)
  • Collaborated in developing a long short-term memory (LSTM) based Kalman filter resulting in ~15% improvement in tracking of vehicles. (Deep learning, Bayesian filtering, python, TensorFlow)
  • Collaborated in developing a convolutional neural network-based visual odometry and ego-localization system resulting in ~10% improvement in localization. (Deep learning, localization, python, TensorFlow)
  • Prototyped an ARM SoC-based surround view system enabling Continental to market a solution 600% cheaper than existing solutions. (Multi-view computer vision, OpenCV, Eigen, C++, embedded system, multi-threading)
  • Enabled extraction of relevant image data from huge datasets by automating image labelling using single shot detector and faster-RCNN. (Deep learning, computer vision, python, tensorflow)

Project MANAS – Manipal Institute of Technology

Manipal, KA, India

Mentor / 1-year | Automation Head / 1-year | Member / 6-months
Sep 2015 – Feb 2018
  • Led a sub-division of 7 undergraduate researchers in the development of:
    • Extended Kalman Filter based Lidar and Radar sensor-fusion system.(Robotics, sensor fusion, ROS, C++)
    • An embedded distributed control and sensor interfacing system. (Robotics, ROS, embedded C, CAN)
    • State lattice based planner for Ackermann steered vehicle. (Robotics, motion planning, ROS, C++)

For our prototype autonomous car propelling Project MANAS to the top 13 finalists nationwide (amongst ~260 participants) in Mahindra Rise Prize Driverless Car Challenge.


  • Secured the first position at Northrop Grumman audio signal processing and classification challenge 2019.
  • INCOSE Associate Systems Engineering Professional (One of 1074 ASEPs worldwide).
  • Student Ambassador at Maryland Robotics Centre, College Park, MD, USA.