MLeap
lastmod 2026-11-10created 2026-07-282 edits
MLeap is a serialization format and execution engine for deploying trained machine learning pipelines to production. It exports pipelines built with Spark ML, scikit-learn, TensorFlow, and others to a common bundle format, and lets single predictions run with low latency on a JVM-based runtime without a Spark cluster. It is used when you want to separate the training environment from the serving environment.
Deploy ML Pipelines to Production