Matt Zinman

Hi 👋. I'm interested in engineering, data journalism, and civic technology. Currently I work as a software engineer at Peloton where I build, design, and maintain Peloton's highly scalable metrics ingestion pipeline and related services like the in-class leaderboard. You can reach me on LinkedIn or by email at mattalexzin@gmail.com.

Curious to see what I look like? Click here

Peloton

Software Engineer

Aug 2022 - present

On Peloton's leaderboard team I work with services that receive more than 2 billion requests per day. Currently I am re-architecting our backend to make our services more resilient to surges in traffic at the beginning of large fitness classes and I am improving the scalability of Peloton's Apple Watch integration. I work with kubernetes, kafka, redis, dynamodb, postgres, argo, datadog, python, kotlin, etc. Some highlights have been:

  • Building and rolling out the micro-service that ingests all of Peloton's leaderboard data.
  • Testing, scaling, and maintaining the leaderboard for Peloton's Ride to Greatness, an annual class where cyclers compete in teams.
  • Making it to semi-finals of the software org's ping pong tournament.

Columbia Journalism School

Research Scholar

Jul 2022 - Aug 2022

I conducted a data-driven investigation into pedestrian safety in Mimai-Dade County, finding that Miami-Dade’s most disadvantaged zip codes are also some of its most deadly for pedestrians.

I also worked on a project to study how political misinformation spreads on Twitter. I analyzed patterns in a giant Twitter dataset that I collected from the Twitter API. I experimented with nlp algorithms to classify tweets, graph algorithms to identify clusters of users, and creative ways of visualizing social networks.

GrassRoutes Canvassing

Co-Founder and Chief Product Officer

Sept 2019 - Nov 2021

GrassRoutes provided local politicians with software tools to run data-driven campaigns at low cost. As the Chief Product Officer I worked with local campaigns to design intuitive and useful product features, and I implemented the frontend of the app using React. GrassRoutes was used by 5+ campaigns in 3+ states during the Novemeber 2020 election. The app was acquired in the fall of 2021.

Social Science Research Council

Jan 2021 - May 2021

I worked as a research assistant for several projects that combined qualitative and quantitative approaches to understand policy's impact on human well-being. You can read my work in A Portrait of California 2021–2022 and A Decade Undone: 2021 Update.

Incite Lab at Columbia University

Jan 2022 - May 2022

I built an interactive webite to visualize the methodology and results of an academic article about geographic disparities in access to mental healthcare.

COMS 4111 Intro to Databases

Teaching Assistant

Sep 2021 - Dec 2021

I mentored students as they completed a project to design a database schema for an application of their choice. Together we thought through how to design schemas for applications like Facebook, or Yelp. When I was originally a student in the class, I designed a database for an application that helped multi-state residents choose which U.S. state they should vote in to maximize impact in national elections.

TheTake.ai

Software Engineering Intern

Jun 2020 - Aug 2020

Have you ever wanted to buy the outfit of a TV or movie character? TheTake lets you shop shows and movies directly from your TV--their API powers LG TV's "Who What Where" feature. As an intern I built a generalized system to scrape data from clothing retailers' websites. The data I collected allowed the company to expand their database of clothes that could be matched to those in TV and movies scenes.

Columbia University Center for Nuclear Studies

Research Fellow

May 2018 - Dec 2018

I developed an interactive lesson plan to help educators teach the geopolitical complications of nuclear proliferation. The lesson plan is used by high school teachers all across NYC, and it is taught in Columbia University's compulsory science class "Frontiers of Science."

I also built webpages to attempt to visually convey the unconscionable scale and scope of nuclear weapons.