Domino is a secure, scalable, and centralized platform for developing, validating, delivering, and monitoring models with full auditability, governance and transparency. It can run in cloud, on-prem, and hybrid environments.
Domino on Kubernetes gives our data science team the flexibility we need while offering peace of mind to our partners in IT.
We can now run Domino across diverse infrastructure, including both on-premises and cloud-based environments, while also plugging into existing monitoring, logging, and enterprise security tools that our IT team knows and trusts.
It's a big step forward on the path to make Data Science a first-class citizen that's aligned with Enterprise IT.
— Joshua Cluff, Data Science Operations Manager, National Oilwell Varco
Domino provides purpose-built functionality across the entire data science lifecycle, from development to production and back.
Domino’s patent-pending Reproducibility Engine automatically versions all project artifacts—code, data, environments, discussions, parameters, and results—so you get complete model lineage, reproducibility, and transparency in a single click.
Gain visibility into every project with a single view of all models, compute costs, environment usage, and model performance.
Domino’s open platform drives high adoption by enabling users to leverage the latest data science tools and packages—open-source and proprietary—while providing transparency and governance.
Data scientists can provision compute resources with a single click, either elastically in a cloud or across on-premises hardware. Domino ships with optimized distributions of the most popular data science tools such as R, Python, Jupyter, RStudio, Tensorflow, and H2O. You can also build your own environment using your tool of choice on Domino’s containerized environments.
Administrators can manage environments, set resource limits, monitor usage, kill run-away jobs, and attribute costs to users or teams.
Domino’s end-to-end platform is a single place for IT, data science teams, and the business to collaborate and manage models across their entire lifecycle.
Domino lets you publish models as REST APIs, as hosted interactive web apps (such as Shiny or Flask), or as scheduled jobs for generating reports or running ETL tasks. This avoids delays caused by re-implementing work in production systems, so the business gets value faster and you can monitor and iterate production models to drive greater impact.
Access controls and gatekeeping features let you enforce governance processes to control deployment.
Analytical assets are stored centrally on managed and monitored infrastructure instead of being spread across users’ machines. Granular access controls keep work secured, with activity logs and reports available to administrators.
Mitigate compliance and audit risk with complete tracking of user activities, secure data access, and full monitoring of models in production.
Domino Control Center provides administrators with total insight into the compute costs of data science work. Compute and spend is accounted for and attributed to users and projects so CIOs and business line leaders can perform more accurate budgeting and planning.