PipelineAI integrates with many model training platforms including Domino Data Labs and H2O.
Consider PipelineAI the "last mile" of your ML and AI pipelines - putting your models into production after they've been trained with DDL or H2O.
You're leaving a lot of runtime optimizations on the table during this sprint to production. PipelineAI gives you the knobs and dashboards needed to fully realize these optimizations.
In addition to performing GPU code generation, 8-bit quantization, and network-layer collapsing, PipelineAI provides optimized model-serving runtimes that are not available in any other product.
We embrace DDL and H2O fully. We can serve any model that they train - and then some!