Domino Manages Data Science Life Cycle

With more and more data available to analyze, organizations are realizing the value that sophisticated analyses using artificial intelligence and machine learning (AI/ML) can provide. The benefits of these analyses are significant: our Dynamic Insights research on machine learning finds that organizations most often benefit through competitive advantage, but also improved customer experiences, increased sales, the ability to respond quickly to opportunities in the market, and lower costs. However, the need for specialized skills to deploy AI/ML models can stall data science initiatives. An organization’s efforts to scale data science and apply models are often complicated by a lack of self-service access to infrastructure, tools and data.

Analytics vendors have made it easier to build and deploy models, embedding AI/ML into many types of applications to reduce the time it takes to create production-ready models. Data science platforms like Domino Data Lab streamline data science operations by providing access to data, integrating quantitative tools, facilitating collaboration and monitoring the development of models.

Often considered in the same context as automation tools designed to help data analysts build models by automating tasks in data science, Domino supports the broader context of centralizing data science work and infrastructure across the enterprise for collaboratively building, training, deploying and managing models faster and more efficiently. System components include:

• Domino Data Science Platform for creation, deployment and maintenance of production models at scale
• Domino Model Monitor for managing the performance of all models across the organization
• Workbench for integrating preferred tools to collaborate on reproducible and reusable models
• Launchpad for exporting production models to any infrastructure and monitoring their impact
• Domino Collaboration for storing all code, data, comments and results in a centralized location so workers can share, review and discuss projects as well as use tagging capabilities to promote idea generation
• Domino Reproducibility Engine for easily locating previous work, reproducing results and providing context-specific recommendations for content that may be useful

Domino is compatible with all major cloud software and on-premises hardware. The unified, open platform enables workers to develop, validate, monitor and deliver models at scale, making data science initiatives more accessible across the organization. Data scientists can better manage the analytics environment, access scalable resources to run complex and multiple tasks in parallel, and share and activate analytic models.

Workers have the advantage of collaboration to speed development since all of the organization’s output is stored in a central analytics hub, making it viewable and reproducible. Models are created and managed by data scientists, not computer engineers, reducing operational costs and increasing analytics productivity. Centralized management of changes across the enterprise means teams can collaborate more effectively and infrastructure governance is more manageable for IT groups. Combined, these benefits could increase an organization’s return on investment by reducing time required to build and deploy models.

Domino Data Lab recently announced an expanded relationship with NVIDIA to increase accelerated computing capabilities in its platform. Domino’s certification for the NVIDIA AI Enterprise will expand market opportunities as Domino Data Lab will run seamlessly on mainstream NVIDIA-certified systems and servers. The collaboration enhances ease-of-access for NVIDIA customers.

Domino employs enterprise-grade security to maintain a secure data science operation, safeguarding against regulatory and operational risks. Data access is governed consistently through permissioning, single sign-on and credential propagation. Data governance resources include tools that maintain a secure history of project changes, creating an auditable environment that will satisfy regulatory requirements.

Expectations are high for data science teams to deliver business impact. Organizations may find that current resources, including people, processes and tools are not sufficient to create and manage sophisticated data science efforts. More than one-quarter (27%) of organizations participating in our Machine Learning Dynamic Insights research said that not having enough skilled resources and difficulty building and maintaining ML systems are pressing challenges. Identifying an appropriate data science platform can help further data management efforts by supporting the workforce and operational aspects of AI/ML as well as model development and deployment.

A data science platform unifies people, tools and work products used across the data science life cycle, developing a culture of collaboration and continuous learning that provides a competitive advantage through the increased value derived from data. Workers need the ability to utilize more tools and spend less time waiting for models and experiments to run, and data science team leaders need information to manage teams and prioritize work more efficiently. For a streamlined approach to managing the data science life cycle, organizations should examine Domino Data Lab’s data science platform for its self-service tools that help accelerate exploration and analytics.