• Collaborative Robotics and Intelligent Systems (CoRIS) Institute
    Graduate Research Assistant Advised by Dr. Kagan Tumer november 2018 - present
    - Develop methods for addressing reward sparsity and credit assignment problems in multiagent settings that require a high degree of inter-agent coordination.
    - Design diversity search methods for improving zero-shot generalization to changes in task dynamics, agent policies and team compositions.

  • C37 Collective
    Applied AI - Artist Researcher july 2022 - present
    - Investigate the application of computational and evolutionary techniques as an aid in creating aleatoric temporal art forms (music).
    - Design tools for multi-modal visualization of the transformation of input as it is subjected to generative models.

  • RedLynx Oy
    Applied AI - Research Engineer july 2019 - september 2019
    - Investigated the confluence of traditional tree-based planning and reinforcement learning for adversarial zero-sum games.
    - Improved Quality-Diversity methods for automated inference of latent spaces which can be used as proxies for a behavior space of policies.

  • Ubisoft Entertainment SA
    AI & Physics Programmer july 2016 - june 2018
    - Implemented policy gradient and evolutionary (NEAT) methods to build an end-to-end pipeline for automated game testing.
    - Designed and developed a web-first rigid body physics engine in TypeScript and Node.js for building online simulators.
    - Designed and implemented a framework for automated focus control, in C++, as a drop-in library for auto-generating user interface mapping across several hardware platforms.
    - Improved the implementation for continuous collision detection in a AA physics engine for performance speed-ups of over 73%. The engine was part of actively supported games with over 10 million+ users.
  • BMC Software
    Software Development Research Intern august 2015 - august 2016
    - Developed an event analysis tool for root cause analysis and mining event associations using a variation of the Rete algorithm for real-time analysis of network events for the network automation team.
    - Implemented a RNN in Python and TensorFlow to predict potential network failures.
  • Pune Institute of Computer Technology
    Undergraduate Research Assistant Advised by Dr. Mukta Takalikar october 2014 - august 2016
    - Compiled data and helped develop methods to improve Named-entity recognition for Hindi and its dialects.
    - Conducted a literature survey on the developments in Neuroevolution techniques specifically for learning in problems with sparse rewards.

academic service

  • + Reviewer for IEEE Transactions on Evolutionary Computation.
  • + PC for the 22nd International Conference on Autonomous Agents and Multiagent Systems (AAMAS; ALA) 2023.
  • + Reviewer for the International Conference on Autonomous Agents and Multiagent Systems (AAMAS) 2021, 2022, 2023.
  • + Reviewer for the Genetic and Evolutionary Computation Conference (GECCO) 2022, 2023.

teaching assistant
+ Machine Learning and Data Mining Spring 15
+ Data Structures and Algorithms Fall 15
+ Programming Paradigms and Generics in C/C++ Fall 14
+ Operating System Administration Spring 13
+ Introduction to Functional Programming with Haskell Fall 13