By Domino Data Lab on February 19, 2020 in Company Updates
This year’s 2020 Gartner Magic Quadrant for Data Science and Machine Learning Platforms is now out, and we at Domino are extremely honored to be named a Visionary for the third time. We earned this honor in the Gartner Magic Quadrant, we feel, because of our product vision and roadmap with intentional focus on the needs of enterprises with large teams of code-first data scientists. As the director of Data Science for one large finance company shared on Gartner Peer insights:
“Their [Domino’s] platform is showing our technology leaders what a modern analytic stack needs to be and has caused real change here.”
It’s always gratifying to hear from both our customers and analysts that we’re on the right track and delivering meaningful value to help companies accelerate their data science work. There’s more pressure than ever on data science teams as organizations recognize the power of data science to create a business advantage. We listened to our customers and doubled down on efforts across product development and customer support, bringing forward a broad range of capabilities to help our customers industrialize data science.
From a product perspective, we believe three aspects of the Domino platform, in particular, are foundational to earning this illustrious moniker: openness, collaboration, and reproducibility.
From the very beginning, we took an open approach, enabling teams to centrally manage hardware and allowing data scientists to use their preferred languages and tools so they can stay on the cutting edge. We’ve really pushed the envelope to help our customers future-proof their data science infrastructures and integrate them seamlessly into their existing stack. Domino now installs on any infrastructure — cloud (including of course Amazon Web Services, Microsoft Azure, and Google Cloud Platform), on-prem, and even in their own Kubernetes clusters. We also rebuilt our whole compute grid to leverage Kubernetes to distribute data science workloads and deploy production models, eliminating the DevOps hassles that data scientists face, reducing the burden on IT and decreasing infrastructure costs.
When data experts can easily share ideas, learn from each other, and build on past work, great things can happen. And when they can more easily collaborate with IT and business stakeholders, they can better ensure they’re building models that matter and more quickly get new models into production. So, we’ve gone even further to make collaboration seamless—both within data science teams and across IT and the business. For example, our Experiment Manager, a “modern lab notebook,” enables data science teams to seamlessly track, organize and manage experiments while our Product Portfolio Dashboard Project enables managers to easily see where projects are in the project lifecycle and identify any that need immediate support.
Collaborative capabilities like these and others are helping our customers beef up their R&D efforts, accelerate deployment processes, and gain all-important visibility into deployed models, in-flight projects, and infrastructure costs. As the chief data scientist at one manufacturing company wrote on Gartner Peer Insights:
“The Domino Data Lab platform streamlines every component of the data science/machine learning workflow from version control, environment management, to reproducibility and final deployment. It has allowed us to deploy machine learning solutions to customers 10x faster than we could have before, and opened up entire new possibilities for collaboration across the corporation.”
The ability to reproduce model results is essential both for rapid innovation and effective governance and management. With reproducibility, data scientists can build on past work to pursue even more ambitious and complex projects. Business stakeholders can see how models are built so they’re more confident in the models and, therefore, more likely to use them. And compliance staff gain transparency into models, including model dependencies and development decisions, so they can better mitigate risk. Our Reproducibility Engine automatically tracks all aspects of data science experiments (code, environment, data, and more) so companies never lose work, can always reproduce their results, and have the transparency they need.
Together, these are a powerful combination, and one our customers repeatedly cite as vital to their success.
We’re truly honored to be recognized for our work in creating an enterprise-strength platform, and are rolling up our sleeves to pack even more functionality into our platform. We have a strong roadmap in place, one we believe will help our customers successfully navigate a rapidly evolving marketplace. We’ll continue to ensure Domino is available on any platform and allows use of any tool or framework. We’ll offer even more capabilities so companies can more easily industrialize data science at scale. And we’ll help them better manage and monitor models once they’re in production so they can better compete in today’s dynamic marketplace.