Model Serving — AI Glossary

The infrastructure and processes for deploying trained AI models to handle real-time or batch prediction requests. Includes load balancing, scaling, versioning, A/B testing, and monitoring. Efficient model serving is critical for production AI systems that need low latency and high availability.