Analytics · ML · Data Engineering · Climate Tech

Kevin Hu

I build data infrastructure and analytics systems that help people understand the world — currently working on climate at Watershed. I also play violin, photograph people and animals, and care deeply about animal welfare.

LinkedInEmail
About

I've worked across most of the data stack — data infrastructure, building AI tooling, and dashboarding in climate tech, predictive models in fintech and logistics, some developer tooling along the way. I like work where I'm supporting my teammates, learning new things, and building something that matters.

Outside of that I've also done veterinary volunteer work — shelter medicine at the SF SPCA, wildfire relief, rescue operations with the Humane Society. Endlessly interesting, compassionate teammates, and ruthlessly discerning end users (especially the cats).

PythonMLdbtBigQueryTerraformGCPSigmaSQLxgboostpandasscikit-learnnumpy
Work
2023 — present
Data Scientist
Watershed
Watershed helps companies measure and reduce their carbon footprint. I started, hired, and grew our analytics team and currently work across most of the stack — core product metrics, dashboards, and C-level regulatory reports on one end; data infrastructure on the other. Recent work includes migrating our BigQuery environment to improve governance, building developer tooling to speed up analytics work in dbt, and building an MCP server and Claude Code skill that connects to our data warehouse, dbt projects, and product codebase so teammates can answer questions about the product and business in plain language.
BigQuerydbtPythonGCPSigmaClaude Code
2022 — 2023
Data Scientist
TruckSmarter
Built ML models to help truck drivers find better loads and spend less on fuel — a pricing model, a recommendation engine, and the infrastructure to run them in production.
PythonMLscikit-learnxgboost
2017 — 2022
Data Scientist
Plaid
Five years building NLP and fraud and security ML at a fintech infrastructure company. Built ACH risk scoring, account takeover detection, and NLP models to make sense of credit card transaction descriptions. Ran experiments that pushed onboarding conversion up 10%+. Grew and mentored the data science team from 1 to 10, and led an internal initiative on ML ethics and responsible model design.
PythonMLNLPSQLxgboost
2016 — 2017
Data Analyst
Operator
Dashboards, customer research, and QA — and also wrote the music for some of their online ads.
PythonSQL
2015 — 2016
Quantitative Developer
Eclipse Trading
Built pricing models for treasury bond and commodity futures. Also ran recruiting, which turned out to involve entirely different skills.
Python
Education
Harvard University
A.B. & S.M. in Applied Mathematics
  • Section lead, Harvard Radcliffe Orchestra
  • Flanker & prop, Harvard Rugby
  • President, Applied Math Society
Beyond work
🎻
Violin
I've performed at venues including the Kennedy Center, Leipzig's Gewandhaus, and Orchestra Hall, in ensembles with Yo-Yo Ma and Joshua Bell, and for President Obama and the UN General Assembly. I've also composed and produced a film score.
📷
Photography
I love photographing people — proposals, company events, and professional profiles for startup founders and teams. Check out my photo gallery for a sample of my work!
📐
Math education
I care a lot about math education and have been involved in it in a few ways — teaching competition math to middle and high schoolers through IDEA MATH, teaching at Art of Problem Solving's online school, writing contest problems for Chicago-area regional competitions, and coaching the Chicago ARML team.
🐾
Animal welfare
Former volunteer on the shelter medicine team at the SF SPCA. With the Humane Society of the US I've led shelter operations after the Camp Fire (250+ animals), and worked cases involving an illegal breeding operation and a dog meat farm. Also done wildfire relief vet work with the Sonoma and Butte Humane Societies.