Deploying ML Models in Production

Deploy your ML models in production in one click in Dataiku with high performance and scalable scoring, deployable on the cloud with Kubernetes.

Deploying ML Models in Production

Deploy to production in one click

  •  Empower analysts and data scientists to deploy models into production in a few clicks.
  •  Data cleaning, enriching, preprocessing, as well as models, are bundled together for simplified scoring pipelines.
  •  Deployed models are versioned, enabling users to deploy new versions, compare them and rollback at anytime.
Deploy to production in one click

Scalability & high availability

  •  Handle large quantities of real-time predictions with queuing, parallelism, and load balancing.
  •  Run multiple scoring nodes for full high availability.
  •  Automatic elastic scaling to handle unexpected traffic surges.
Scalability & high availability

Deploy on the cloud with Kubernetes

  •  Deploy your API on-premises or in the cloud.
  •  Fully native integration of Kubernetes for elastic and reproducible deployments.
  •  Full GPU support for deep-learning models.
Deploy on the cloud with Kubernetes

Powerful API engine for your applications

  •  Deploy as an API: visual models, custom Python or R models, custom Python or R functions or SQL queries.
  •  Easy to use REST API. Embed in a few lines of code.
  •  Automatic generation of ready-to-use code samples.
Powerful API engine for your applications

Avoid model drift with a feedback loop

  •  Run multiple versions of the same model at the same time for automated A/B testing.
  •  Monitor data changes over time.
  •  Access history of logs queries and predictions at any time to check that model performance is not drifting with time.
Avoid model drift with a feedback loop

Get Started with Dataiku

Start an online hosted trial, download the free edition,
or compare the features of the Lite, Team, and Enterprise editions.

Get Started