Advanced Analytics Tool Vendor Selection

In an effort to enhance our data and analytics capabilities, and as part of the data & analytics strategy and roadmap, Northeastern University is selecting an AutoML (automated machine learning) tool. A team of Northeastern analysts and leaders, in conjunction with our longtime D&A partners, Eckerson Group, has embarked on a vendor selection project to choose an AutoML tool for Northeastern.

Eckerson kicked off the project by interviewing analysts across the university to understand their needs for advanced statistical modeling, including use cases, types of models, and expectations for available functions and level of sophistication.  The requirements generated by these interviews were consolidated to provide a profile of our needs and a set of questions to potential vendors.  Using Eckerson’s expertise and tools like the Gartner magic quadrant, the project team identified a broad set of potential vendors.

While the initial list of vendors was wide, some did not meet our criteria, some were too expensive, and others didn’t have the time to work with us.  We have now narrowed the list to DataRobot, Microsoft Azure ML, Amazon SageMaker, Aible.ai, and RapidMiner.  These vendors have provided responses to our Request for Information (RFI) and agreed to deliver a 2.5 hour demonstration for our committee of analysts.  The committee is scoring the vendors based on our list of requirements, and these scores as well as qualitative commentary, cost, and ongoing discussion with our group of experts will help us make a final decision by the winter break.

This is an exciting project and we look forward to implementing a tool that will enable Northeastern analysts to create accurate, responsive, maintainable, and ethical models for better decision-making.  For more information, please reach out to Rana Glasgal or Brian McGrath.