Experience

Graduate Work experience

• Summer 2022
  • Microsoft: Research Intern
  • Focus: Interactive Grounded Language Learning
  • Supervisor: Julia Kiseleva
  • Details: Built an interactive learning solution for a challenging multi-modal NeurIPS Competition Task
    • Our framework enables an agent to receive “help” feedback from a human operator when it makes a mistake. The agent also learns how to ask a relevant clarifying question.
    • In submission to ARR.
• Summer 2021
  • Amazon: Applied Science Intern
  • Focus: Interactive Product-Image Retrieval
  • Supervisor: Emre Barut
  • Details: Designed a graph-based, interpretable model for the task of multi-modal image retrieval: Given an image and an utterance describing a modification to that image, retrieve the target image from the database that has that modification
    • Approach achieved State of the Art performance on two challenging publicly available datasets (Fashion-IQ by 1%, Fashion200K by 5%)
• Summer 2020
  • Amazon: Applied Science Intern
  • Focus: Named Entity Recognition
  • Supervisor: Michael Kayser
  • Details: Improved Named Entity Recognition on Alexa utterances with multiple entities by > 10% using gazetteers with neural NER models. Paper published at Amazon ML Conference

Undergraduate Work experience

• Summer 2019
  • Fujitsu Laboratories of America: ML Research Intern
  • Focus: Unsupervised Joke Slection, Given a Dialogue
  • Details: Created a patented, unsupervised approach for selecting a joke to show users depending on their dialogue history. Patent approved here
• Summer 2018
  • Loom.ai: Deep Learning Intern
  • Focus: Computer Vision
  • Details: Used Computer Vision techniques (CNN) to improve facial attribute classification by 18%. Loom.ai was a startup to make 3D avatars from a single facial scan, now acquired by Roblox.
• Summer 2017, 2016
  • UnifyID: SDE, Machine Learning Intern
  • Focus: Machine Learning for User Authentication
  • Details: UnifyID was a startup to authenticate users based on implicit signal data from their phone, such as the way the walk. They were acquired by Prove. As a Machine Learning Intern, Implemented a system to identify users based on how they pick up their phone. As a Software Intern, I improved their mobile applications.