Abstract: No more science experiments with months of data engineering and months of experimentation that produce minimal (if any) business value while simultaneously tying up your best data scientist for a full year. Executives and data science leaders that are frustrated with the pace of delivery for machine learning use cases or who are constrained by labor availability, need to witness the massive (and rapid) impact that is possible with extreme automation in the ML development and ML deployment stages. During this session you’ll see a DataRobot model factory that produces 100s of fully specified, deployable machine learning pipeline, selects the top 10 and deploys them each with a single command. Don't fear that extreme automation requires sacrificing model quality. DataRobot full autopilot models regularly outperform meticulously hand-coded models in an apples-to-apples bake off -- in hours versus months of effort.
Bio: Andrea has spent 20+ years as an analytics professional as the world has transitioned from spreadsheets, to BI tools, to machine learning and AI. Before DataRobot most of her applied experience is with HR use cases especially recruiting, retention, engagement, learning & development, and compensation planning. Andrea also spent 5 years helping small health care businesses with everything related to marketing funnel optimization across organic search, pay-per-click, website design, on-page conversions and retention marketing. Her personal interests include e-commerce and sports analytics.