Liam Hinzman
Software Engineer at Exa
San Francisco, California, United States
Summary
🤩
Rockstar🎓
Top SchoolLiam Hinzman is a software engineer based in San Francisco with 10 years of experience building full-stack systems, ML tools, and multi-agent AI products. Currently leading design and full-stack for Exa Websets, he architects distributed AI deployments that orchestrate thousands of agents to deliver precise web results. His background spans research-grade ML at Numerai—where he helped launch a high-return fund and developed Signals and research frameworks—to production ETL, NLP inference pipelines, and real-time systems. Liam brings a rare blend of product-facing design, backend engineering (Python, Elixir, Postgres, GraphQL), and hands-on model work, having also contributed data-science scripts to Numerai’s example repo. He pairs creative 3D freelance experience with pragmatic engineering, reflecting a cross-disciplinary approach to UX and system design. Known for debugging tricky ML issues (e.g., XGBoost float errors) and shipping reproducible pipelines, he thrives on turning complex data problems into reliable products.
10 years of coding experience
3 years of employment as a software developer
Computer Science, Computer Science at Minerva University
Stackoverflow
1reputation
0reached
1answer
0questions
Github Skills (12)
pandas10
machine-learning10
python10
data-science10
data-analysis10
xgboost9
jupyter-notebook8
scikit8
scikit-learn8
javascript6
jquery6
cryptocurrency4
Programming languages (14)
Github contributions (5)
numerai/example-scripts
Dec 2020 - Jan 2022
A collection of scripts and notebooks to help you get started quickly.
Role in this project:
Data Scientist
Data ScientistContributions:28 reviews, 34 commits, 30 PRs in 1 year
Contributions summary:Liam primarily contributed to code related to data analysis and model training within the Numerai ecosystem. Their work involved modifying existing scripts to align with data format changes, including renaming target columns and standardizing data handling. They also addressed debugging issues related to XGBoost floating-point errors and updating the example model to use a 20-day target.
pythonsciencenumeraidata-sciencemachine-learning
LiamHz/liamhz.github.io
Nov 2018 - Feb 2025
personal site written with vanilla christian html and css
Contributions:7 PRs, 326 pushes, 5 branches in 6 years 3 months
reactcssnextjs
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.