Technical Talks

Jonathan Jin

Staff Machine Learning Engineer | Hinge

Trimming the Long Tail of Production Model Ownership at Hinge

ABOUT THE TALK

ML OPs & Platforms

Beyond model performance lies a critical challenge in machine learning: comprehensive model ownership. This talk examines how focusing on the often-overlooked "long tail" of machine learning infrastructure can dramatically improve operational efficiency and innovation. Staff Engineer Jonathan Jin from Hinge's AI Platform team will reveal how addressing challenges like observability, feature access, and model refinement creates a "golden path" that empowers teams to continuously innovate. Attendees will learn how strategic infrastructure development can transform machine learning from a performance-driven to a holistic, sustainable practice.

Staff Machine Learning Engineer

Jonathan Jin

Hinge

Jonathan Jin is a Staff Machine Learning Engineer at Hinge where, as a member of the AI Platform Core team, he helps democratize the cost-efficient, robust, and scalable use of AI within the company. He has previously worked on production-scale ML platform and infrastructure at Spotify, NVIDIA, and Twitter. Prior to diving down the ML rabbit-hole, he worked on observability and SRE infrastructure at Uber. When he's not debugging inference pipelines, he enjoys cooking, literature, and plot-heavy JRPGs.

Technical Talks

Trimming the Long Tail of Production Model Ownership at Hinge

ABOUT THE TALK ML OPs & Platforms

ABOUT THE TALK

ML OPs & Platforms