Bex Reilly

Product Manager / Data Scientist
San Francisco, CA
rnr.reilly@gmail.com

Education

University of San Francisco

MS, Data Science
June 2019

Boston University

BA, Economics and Psychology
May 2015

Experience

Product Manager

Shipt
January 2022 - Present

Machine Learning Product Manager, Grammar Coach

Dictionary.com
  • Defined and executed product roadmap
  • Studied market trends and user feedback to understand customer needs, and presented insights to stakeholders
  • Collaborated with a cross-functional team to prioritize UI, backend infrastructure, and Machine Learning development
  • Led creation of new features, including plagiarism detection, and inclusive language suggestions; monthly recurring revenue has increased 40%
  • July 2021 - January 2022

    Data Scientist

    Dictionary.com
  • Designed, built, and maintained backend API, using FastAPI, for the Thesaurus.com Grammar Coach product
  • Implemented Natural Language Processing (NLP) techniques such as tokenization, part-of-speech tagging, inflection, and sentiment analysis in various API endpoints using Python and the spaCy library in order to support tone detection and context aware synonym suggestions
  • Leveraged pre-trained machine learning models as a service for the task of Grammar Error Correction, used to serve grammar corrections to users
  • Measured product performance, using F0.5 score, against Grammar Coach competitors through creation of a benchmark dataset, which includes collaboration with the Editorial team and a Mechanical Turk data pipeline, used in combination with an error annotation tool
  • Monitored Grammar Coach logs from the API and 4 service deployments via Elasticsearch, Kubernetes, and Datadog across staging and production to identify areas for improvement in the codebase
  • Partnered with Product and Engineering to design and propose new features, such as sentiment score and contraction expansions, based on NLP capabilities and current literature
  • July 2019 - Present

    Data Science Intern

    Dictionary.com
  • Improved advertisement pricing bucket logistic regression precision and recall by 1% each through discovery of impactful features based on impression-level data
  • Utilized MongoDB to obtain example sentence data and apply NLP techniques such as word similarity and sentiment analysis to optimize classification of sentence appropriateness
  • Initiated idea for Collaborative Filtering based recommendation system using word2vec and Jaccard Similarity in 1st place Hackathon project that would allow users to see which relevant words similar users have searched
  • November 2018 - July 2019

    Data Analyst

    Boston Children's Hospital
  • Identified gaps in enterprise’s tracking of the ‘boarder population’ (stuck patients that use 2% of the institution’s beds) and launched a monthly data extraction and analysis via MicroStrategy for presentation to senior department and hospital leadership
  • Created a monthly audit report on electronic medical record (EMR) documentation of patient risk for harm to self and others for Child and Adolescent Psychiatry Fellows; compliance increased by 15%
  • December 2016 - June 2018

    Research Data Coordinator, Psychiatry

    Boston Children's Hospital
  • Collected, analyzed, and displayed key metrics for senior leadership via reports, presentations, and dashboards across 7 service areas with more than 4,000 patients annually
  • Executed complex queries in large relational data warehouse
  • July 2015 - December 2016

    Skills

    • Python
    • SQL
    • git
    • AWS (ec2, Athena, Kubernetes)
    • Web Servers (FastAPI, Flask)
    • Natural Language Processing (SpaCy, NLTK)
    • Machine Learning (pandas, scikit-learn, NumPy)
    • Visualization (Tableau, Seaborn, Matplotlib)
    • Test Driven Development / Behavior Driven Development

    Continuing Education

    • Certified Scrum Product Owner, Scrum Alliance
    • Mastering Product Management, Reforge
    • AB Testing, Udacity

    Projects

    • Analyzed the impact of bike-sharing ridership on air quality in NYC using Spark, MongoDB, and machine learning techniques (Elastic Net, Random Forest)
      • published in IEEE Smart World Congress 2019