Models as Microservices, Platforms as Partners - Collaboratively Building ML Infrastructure at Hinge

Learn how to build an online model serving platform that ML engineers actually want to use in this 30-minute conference talk from KubeCon + CloudNativeCon. Discover how Hinge's AI Platform Core team designed a self-serve platform for deploying and monitoring online models using Ray Serve, MLflow, Grafana, and other Kubernetes-native tools, while smoothing rough edges with interfaces aligned to internal conventions like gRPC and OpenTelemetry. Explore the platform architecture that enables ML engineers to deploy models as microservices without touching Kubernetes or Helm directly, while providing fine-grained observability and standardized Service Level Indicators (SLIs). Understand how the team achieved over 40% reduction in model production timelines through early partnerships and developer experience focus, fostering a culture of collaboration and trust that drove broader adoption across Hinge's engineering organization. Gain insights into the strategic approach of treating models as microservices and platforms as partners in building effective ML infrastructure that balances technical sophistication with usability.