Zoe Weil

Zoe Weil

About

Detail

Co-Founder/CEO
New York, United States

Timeline


work
Job
folder
Project
auto_stories
Publication

Résumé


Jobs verified_user 0% verified
  • i Members
    Member
    i Members
    Apr 2024 - Current (2 years 4 months)
    3i brings engaged private investors together to source opportunities, share expertise and build value for one another that goes far beyond the deal.
  • Sequen
    Co-Founder/CEO
    Sequen
    Jan 2024 - Current (2 years 7 months)
    Were hiring!!! https://www.sequen.ai/careers The first Behavior Design Engine for the enterprise. Sequen isn’t retrofitted AI search or recommendations. It rethinks relevance from first principles. Sequen introduces the first foundational Large Event Model (LEM), trained on billions of user event sequences and built natively on a reinforcement learning infrastructure. LEMs are specialized neural networks that predict the next user event—just as LLMs predict the next word. Sequen’s LEMs are pre-trained on billions of user-site interactions and fine-tuned to optimize for the outcomes you care about. No more fixed pipelines with fragmented infrastructure. Sequen replaces them with a single endpoint that adaptively handles all phases of pe
  • Angel Investor
    Angel Investor
    Angel Investor
    Jan 2024 - Current (2 years 7 months)
    Personal Investor/Shareholder in: Inboundr AI Hyperparam Tough Day Eden Labs Starcycle Adaption Polymathic
  • Blue Tulip Ventures
    LP | Advisor
    Blue Tulip Ventures
    Jan 2024 - Mar 2024 (3 months)
    Advisor at Blue Tulip Ventures, guiding investments in virtual humans and AI clones to shape a future where technology augments human potential.
  • Etsy
    Staff Applied Scientist
    Etsy
    Aug 2020 - Dec 2023 (3 years 5 months)
    * Led overall ML strategy for Etsy’s QU initiative; pioneered work on adoption of SOTA LM/ Transformers; Led AML Labs NLP reading group * Query Labeler – Led the development of Query Classification platform powering Broad/Direct query classification + mature query classification * Lingo –Led the design/research/development/implementation/launch of Lingo, the first intelligent Query Suggestion system at Etsy and first LLM fine-tuned on Etsy data. Lingo also served as a proven foundational blueprint to other projects from other teams including SM and Recsys + wins * TGen – Led design, architecture generative model/platform for user/listing/query feature generation this project also contributed 4 of the most powerful features to Etsy’s first
  • Risalto Health
    Machine Learning Advisor
    Risalto Health
    Jul 2020 - Jan 2023 (2 years 7 months)
  • Mindful Machines
    Co-Founder
    Mindful Machines
    Oct 2018 - Aug 2020 (1 year 11 months)
    Invited to speak at several international conferences throughout the world including:
  • Risalto Health
    Chief Scientist — Artificial Intelligence
    Risalto Health
    Mar 2018 - Jul 2020 (2 years 5 months)
    • Build an entire deep learning platform from scratch using a combination of custom and open-source libraries with Scala, MxNet, Sockeye • Implemented a series of state-of-the-art, deep transformer/LLM models (BERT, roBERTa, Reformer) from research papers into testing and later as a full-suite product to predict a patient’s risk requiring future medical treatments such as surgery. • Design, architect, deploy, an end-to-end reporting platform using python, pweave, scikit-learn, numpy, and airflow for rapid research and testing of business hypothesis • Built DevOps infrastructure in terraform for running and scaling ML/DL on AWS
  • D
    Data and Semantic Analysis - Tech Lead
    DRESR Inc bought by Google
    Sep 2016 - Mar 2018 (1 year 7 months)
    • Architect, design, build entire Healthcare Data Pipeline end-to-end using AWS, RabbitMQ, Scala, and Postgres • Design and implement security layers for data pipeline using Vault • Introduce the use of a SemanticDB in conjunction with new-style Scala Macros to remove boilerplate and reduce the size of the codebase • Architect streaming context-based recommendations using kafka
  • Komodo Health
    Head Engineer, Machine Intelligence
    Komodo Health
    Jul 2015 - Jul 2016 (1 year 1 month)
    • Architect, design, build an intelligent and adaptive learning system that predicts and improves patient adherence to medication (Scala, Spark, Airflow) • Architect, design, build an Entity Resolution system to resolve relationships in a universe of doctors using public and proprietary data (Scala, Spark, Kubernetes, Airflow)
  • HookLogic Inc Acquired by Criteo in
    Lead Machine Intelligence Engineer
    HookLogic Inc Acquired by Criteo in
    Oct 2013 - Jul 2015 (1 year 10 months)
    • Build and lead a team of five Data Scientists • Devolop, meta-optimize and implement machine learning algorithms (Hadoop/Hive/Scala) to predict Conversion Rate and Click-Through Rate across advertisers/publishers • Utilize Akka Actors and message-based concurrency to reduce costs of computation while training and testing data over large datasets (scala) • Determine equations for significance testing • Deploy live machine learning quality score models and analyze their performance (Scala, Java EE, Oozie, AWS) • Perform A|B tests on live models • Re-write feature-building code for our machine learning framework to help us better understand and dissect model performance • Architect, design, and build an end-to-end taxonomy recommendation eng
Education verified_user 0% verified
  • N
    Doctor of Philosophy - PhD, Mathematics and Computer Science
    NYU Courant Institute of Mathematical Sciences
Projects (professional or personal) verified_user 0% verified
  • R
    Recommender System
    May 2014 - Aug 2014 (4 months)
    1. Create a recommender system which combines content-based filtering with collaborative filtering- implemented using Java and Hadoop - dramatically extended Apache Mahout source code in the process. 2. Fix a Mahout issue where using a Boolean data set to calculate Euclidean distance was impossible. To counter this issue, we took advantage of the concept of “anomalous co-occurrence” and calculated LLR's. We were also able to improve the root mean square error (RMSE) of recommendations by more than two orders compared with the original Boolean data
  • P
    Peapod
    Peapod is a dependency and data pipeline management framework for Spark and Scala. The goal is to provide a framework that is simple to use, automatically saves/loads the output of tasks, and provides support for versioning. It is a work in progress and still very much experimental so new versions may introduce breaking changes.
  • L
    LadyDi
    The goal of LadyDi is to help you Code Less, Build More. It is very easy to use and provides users with clean, automated Feature Generation and Selection for Apache Spark. It is a work in progress and still very much experimental so new versions may introduce breaking changes.
Publications verified_user 0% verified
  • a
    adSformers: Personalization from Short-Term Sequences and Diversity of Representations in Etsy Ads.
    Feb 2023
  • H
    Online, Continuous Hyper-parameter Optimization in Industry Using Spark
    HookLogic
    Feb 2015
This is a community-created genome.