Infrastructure Foundation in the Domino platform

Remove DevOps pain with one-click unlimited compute and parallel experimentation

Domino provides a technology foundation to meet the unique needs of experimentation and model development. Instantly run code on machines with up to 2TB of RAM and 100 cores, or on specialized GPUs for deep learning; get all the power of Docker and containerization without being a DevOps expert.

Run multiple experiments in parallel to iterate faster, while Domino’s patent-pending Reproducibility Engine automatically tracks and organizes all artifacts and results of each experiment.

Accelerate research with your choice of innovative tools

Domino comes pre-configured with optimized computing stacks for research. They include popular languages, tools, and packages such as R, Python, SAS, Jupyter, RStudio, Tensorflow, and H2O. Or take advantage of new data science tools by building your own custom environment with any packages you’d like.

Environments can be shared among your colleagues but safely sandboxed from others.

Unlike other platforms that force you to run code in their proprietary notebook, Domino lets you run code in a native development environment such as RStudio, SAS Studio, or JupyterLab.

Domino Open Compute Environments
Knowledge Center in the Domino platform

Eliminate re-work and painful reconstruction of experiments

Domino provides a single place for you and your teammates to build, validate, deliver, and monitor models. The Reproducibility Engine automatically tracks all experiments, so you never lose work and can always reproduce results. Domino keeps code, environment details, data, discussions, parameters and results for each experimental step.

Data science is a research-based process and Domino allows you to find, understand, build on, and contribute back to the collective expertise of your organization. Use existing code snippets, pipelines, or other artifacts to reduce redundant tasks such as manual data cleansing.

Reduce deployment friction with self-service model delivery

Domino supports multiple modes of model delivery, so that models fit seamlessly into existing workflows and systems to maximize impact. Stakeholders can receive scheduled reports, use self-service web forms, and engage with full-fledged apps built in Shiny or Flask.

IT engineers of downstream systems can call batch and real-time APIs which are automatically versioned, secured, and highly available. Domino automatically tracks the full provenance—code, environments, data, parameters, and discussion—of a delivered model so that IT can validate, approve, or debug.

Model delivery in the Domino platform

Spotlight for Data Scientists