More Cash for DataRobot Along with ML Ops Tool
High-flying enterprise AI specialist DataRobot announced another huge funding round along with a machine learning platform for managing predictive models that combines internally developed monitoring framework with so-called ML Ops tools acquired earlier this year.
Boston-based DataRobot said Tuesday (Sept. 17) it has added another $206 million to its venture capital war chest, bringing its investment total through seven funding rounds to $431 million. It announced a $100 million Series D funding round last October.
Alliance Bernstein PCI, EDBI, Tiger Global Management and World Innovation Lab joined the Series E funding round as new investors, joining existing investors that include Intel Capital.
The latest cash infusion will be used to expand development of DataRobot’s new ML Ops framework along with the automated machine learning and times series components of its flagship platform.
Earlier funding was used for several acquisitions, including the June deal to buy ParallelM, the machine learning operations specialist. Those ML Ops tools are used to scale deployment, management and governance of models in production settings.
The ML Ops initiative announced this week by DataRobot is designed as a “hub” for deploying and tracking of predictive models created using a range of tools. Among the goals is addressing the relatively small number of AI models currently moved to production. Among the hurdles is a lack of governance and monitoring required to ensure trusted AI models.
DataRobot said ParallelM’s technology deploys and manages machine learning models built on different machine learning platforms onto customer-managed environments, including Kubernetes and Spark. The combination is touted as providing real-time monitoring and centralized management of models on an open platform.
Ultimately, the integration of ML Ops capabilities into DataRobot’s machine learning platform aims to automate AI deployment as a means of pushing more trusted models to production. So far, investors are onboard with its approach.
Underpinning DataRobot’s approach are open source algorithms in R, Spark ML and other languages. The automation platform pits those algorithms against each other to determine which works best. The optimum machine learning models are then deployed in production on a variety of big data clusters.
“Machine learning operations and governance is a must-have to become an AI-driven enterprise,” DataRobot CEO Jeremy Achin asserted in announcing the ParallelM acquisition. That deal was the seven-year-old company’s fourth in the last two years.
DataRobot joins a growing list of enterprise vendors targeting the management of machine learning deployments. For example, Hewlett Packard Enterprise (NYSE: HPE) last week announced its entry into the growing ML Ops market with a suite of software and tools for managing models. Those tools combine container-based software for AI development and management acquired in a deal last year for BlueData.