Jonathan Jin
Machine Learning Software Engineer
About Me
I’m a machine learning engineer with a specialization in platform and infrastructure.
Most recently, I was a senior member of Spotify’s ML Platform org. There, I worked on:
- Compute orchestration and workload management, e.g. managed multi-tenant cluster infrastructure for Kubeflow and Ray via Kubernetes in Google Cloud;
- Developer-friendly AI/ML governance and artifact/experiment management with Backstage and MLflow.
Before Spotify, I helped build machine learning platforms at NVIDIA as a member of their autonomous vehicles division and Twitter as a member of their Cortex organization. I’ve also worked on site reliability and observability infrastructure at Uber.
I’m a proud alumnus of the University of Chicago, where I studied computer science and economics.
Resume
Talks
- “How Spotify is Navigating an Evolving ML Landscape with Hendrix Platform”, TWIMLcon (2022)
- “Empowering Traceable and Auditable ML in Production at Spotify with Hendrix”, MLconf San Francisco (2022)
- “Scaling Kubeflow for Multi-tenancy at Spotify”, KubeCon + CloudNativeCon North America (2021)
- “ML Workflows at Twitter: Lessons Learned”, AI NEXTCon New York (2019)
Contact
About This Site
This site is written with Emacs as a standard Org mode file, exported to static HTML with Water.css for styling. It uses PDF.js for in-line resume display on browsers that support it.
This site tracks site visits using Fathom, a privacy-focused Google Analytics alternative that collects no personal data.
The site is hosted on GitHub with GitHub Pages. Take a look at the (truly very tiny) repo here, if interested.