Experience

• 

  Gantry

  • San Francisco, California, United States

  • Machine Learning Engineer

    • Jan 2023 - Jan 2024
    • San Francisco, California, United States

    - Built an LLM evaluation workflow using existing LLMs to measure output quality along varied performance metrics- Revamped internal LLM prompts to improving LLM output robustness and completely eliminating hallucinations during model evaluations.- Added product features to report metrics on "classical" ML models, such as F1-scores, KL-divergences, etc.- Contributed towards R&D efforts towards automatic hallucination detection models for LLMs

• 

  Fiddler

  • Palo Alto, California, United States

  • Machine Learning Engineer

    • Feb 2021 - Dec 2022
    • Palo Alto, California, United States

    - Developed new XAI techniques for production ML settings.- Rebuilt model monitoring pipelines to reduce metric computation latency and increase data throughput.

• 

  Berkman Klein Center for Internet & Society at Harvard University

  • Cambridge, Massachusetts, United States

  • Research Scholar

    • Jun 2021 - Aug 2021
    • Cambridge, Massachusetts, United States

    Worked with Saidot and the City of Helsinki to develop a framework to deploy AI projects ethically and responsibly in public sector projects.

• 

  Harvard Business School

  • Cambridge, Massachusetts, United States

  • Research Assistant

    • Jan 2020 - Mar 2021
    • Cambridge, Massachusetts, United States

    - Conducting explainable machine learning research with Prof. Hima Lakkaraju at the AI4LIFE (Artificial Intelligence for Large-scale, Interactive, Fair, and Explainable decision-making) research group.- Project topics include counterfactual explanations, recourse, and interactive human-in-the-loop artificial intelligence.

• 

  Nanyang Technological University

  • Singapore

  • Research Intern

    • Aug 2018 - Dec 2018
    • Singapore

    - Wrote social network analysis based thesis "Co-operative Influence Maximization via Contrasting Opinion Formation", under the NTU India Connect program.- Paper accepted for publication at the IEEE BigData conference, 2019.

• 

  Institute for Artificial Intelligence, University of Bremen

  • Google Summer of Code

  • Software Developer

    • May 2018 - Aug 2018
    • Google Summer of Code

    - Rewrote portions of the PracMLN machine learning software package. - Improved efficiency in learning and inference using Markov Logic Networks.- Produced an overall speed increase of ~30%.

• 

  SENSEI Technologies

  • Bengaluru Area, India

  • Software Development Intern

    • May 2017 - Jul 2017
    • Bengaluru Area, India

    - Created an interactive neural-network powered product, deployed via an API on a Flask server.- Expanded the company tech-stack to include deep learning frameworks, allowing for integration of AI technology into software development projects.