Scaling AI at Skyscanner with Mosaic Model Serving

Deploying AI in production is getting more complex — with different model types, tighter timelines, and growing infrastructure demands. In this session, we’ll walk through how Mosaic AI Model Serving helps teams deploy and scale both traditional ML and generative AI models efficiently, with built-in monitoring and governance.We’ll also hear from Skyscanner on how they’ve integrated AI into their products, scaled to 100+ production endpoints, and built the processes and team structures to support AI at scale.

Key Takeaways:

• How Skyscanner ships and operates AI in real-world products
• How to deploy and scale a variety of models with low latency and minimal overhead
• Building compound AI systems using models, feature stores, and vector search
• Monitoring, debugging, and governing production workloads