Self Service Data Science
Domino's Enterprise MLOps Platform let you run experiments on powerful, on-demand compute resources and create production-ready models in seconds using the same tools and languages you love
Self-serve elastic compute
One-click access to scalable compute
Say goodbye to DevOps learning curves and wait times. Stop trying to guess compute needs in advance. With Domino, you can self-serve dynamically adjusting Kubernetes-based compute clusters with just a few clicks. You can easily access distributed frameworks such as Spark, Ray 2.0, and Dask, as well as NVIDIA GPUs, to power the most computationally hungry algorithms.
Software and environment management
The right tools for the job, centrally provisioned
Modern data science teams use dozens of tools and multiple languages every day. Experimenting with new tools is critical, but can be a nightmare to reproduce. This limits innovation and hurts collaboration.
Domino provides one-click access to a wide variety of open-source and commercial tools and languages including Python, R, SAS, and MATLAB. Versions are tracked to avoid conflicts when team members try to reuse work. As technology changes, it's easy to add new emerging tools when needed, future-proofing your data science platform.
Unified data access layer
Easy, governed, and secure access to data
Data is the fuel for all models but data access and preparation is a regular struggle. Not with Domino.
Domino’s secure Data Connectors provide rapid, secure access so you can quickly get to work while also adhering to data access policies. Domino’s Data Access Library unifies access patterns for disparate data types through SQL syntax. Analytic-ready data, along with its associated metadata, is easily saved in a result set for later reuse, saving time and compute costs.
Out-of-the-box support for Feast– the emerging, open-source standard for feature stores – increases the reproducibility, consistency, and reusability of features while also helping to mitigate skew between the features used in training and those used in production. Organizations now have a single source of truth for calculating important features.
Audit trails and granular data access controls ensure that only team members authorized to see data can do so, and you can show how access has changed over time.
model and app deployment
Easily export production models to infrastructure of choice
Building models is tough enough — deploying them shouldn’t be. Don't waste time translating results into other languages.
With Domino, models are easily deployed through scalable APIs, Apps, and Launchers, or exported as Docker images to CI/CD pipelines, AWS Sagemaker, or other infrastructure. Interactive, scalable Apps created with Shiny, Dash, and Flask make it easy for non-technical users to interact with models.
Self-Service data science resources
Pump Up Data Science Productivity With a Modern Workbench
Spark, Dask, and Ray: Choosing the Right Framework
Tensorflow, PyTorch or Keras for Deep Learning