SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Patreon is seeking a Senior Machine Learning Engineer, Infrastructure to join the Relevance team, which builds ML systems powering creator discovery and content surfacing across the platform. The role focuses on architecting and scaling high-throughput, low-latency live inference infrastructure that serves millions of creators and fans.
Key responsibilities include:
- Architect, scale, and maintain production-grade ML inference infrastructure supporting search, feed ranking, and creator-fan matching systems
- Own the end-to-end feature store lifecycle, from data ingestion and transformation through production serving, ensuring high availability and consistency between online and offline features
- Design and implement comprehensive observability, monitoring, and validation frameworks to detect performance gaps, latency spikes, and production drift
- Collaborate cross-functionally with Product, Data Engineering, and Trust & Safety teams to translate product requirements into scalable infrastructure solutions
- Automate model deployment and reliability testing to improve developer velocity and system stability
- Debug complex relevance systems when monitoring identifies performance bottlenecks or reliability issues
Required qualifications:
- Deep experience building, deploying, and maintaining production-grade ML infrastructure at scale, specifically with low-latency live inference pipelines and feature store architectures
- Strong background in distributed systems and backend engineering with robust, maintainable code in Python
- Systematic approach to debugging complex, high-throughput systems and performance bottlenecks
- Passion for building foundational '0 to 1' infrastructure systems that provide reliable foundations for teams
- Strong communication skills and ability to create clear documentation for system architectures
- Growth mindset with attention to detail in code reviews and commitment to empowering teammates
Paratreon powers over 300,000 creators with $10 billion+ generated since inception and 25 million+ paid memberships. The role is hybrid (3 days/week in-office) based in San Francisco or New York.
