Sawyer Birnbaum

Wharton MBA | Stanford CS Master's | Developing next-gen LLMs

Palo Alto, California, United States

About

Experience

  • Inception (San Francisco Bay Area)
    • Chief of Staff
      Sep 2025 - Present · 11 mos

    • Business Lead
      Sep 2024 - Sep 2025 · 1 yr 1 mo

  • Intern at Anthos Capital
    Jun 2023 - Sep 2024 · 1 yr 4 mos

    • Designed data-driven sourcing system and led team developing system; system has identified dozens of top-of-funnel prospects • Established methodology for evaluating alternative data providers and ran user trials to reach purchasing decisions • Developed processes and tools for analyzing firm's pipeline and portfolio to inform investment decisions

  • Machine Learning Engineer, Product Manager, and Data Scientist at Afresh
    Jun 2019 - Aug 2022 · 3 yrs 3 mos

    • Led ML team efforts to quantify and monitor performance: e.g., designed metrics and built tools to track Afresh’s reduction of waste and to detect and diagnose software and data quality issues • Managed projects including (a) a system that balances weekly orders to meet distribution center capacity limits; (b) a refactoring of Afresh’s codebase that increased efficiency by 15%; and (c) a system that tracks inventory levels for produce items based on sales and shipments • Built deep learning model predicting demand for fresh produce, improving performance by 20% • Interviewed candidates for engineering roles, including leadership positions on the ML team • Developed company-wide IT security policies and procedures needed for SOC 2 compliance

  • Program Manager Intern at Microsoft
    Jun 2018 - Sep 2018 · 4 mos

    • Developed model to classify the source of build failures; increased accuracy from 20% to 98.8% • Created system for triaging failures based on their classification; used by thousands of Microsoft developers

  • Software Engineer Intern at Facebook
    Jun 2017 - Sep 2017 · 4 mos

    • Demonstrated effectiveness of a signal (i.e., a user action) as a proxy for ad conversions using MySQL- and R-based data analysis; integrated the signal into the ads targeting system • Trained machine-learning models to compare the predictive strength of signals