Agile Experimentation – from ideas to deployment

Abstract: As investment in data science continued to grow at CCC, the technology department adapted the software development lifecycle. From idea generation that collaborations between technology, strategy and the business, to quick experiments that test those ideas against the baseline, to an enhanced QA process that measures accuracy not just functionality, to deployment on a hyper-scale platform and finally to the support processes that make sure the predictions are within spec and a way to re-train and improve the models, this new software development lifecycle is at the forefront of process innovation.

Bio: John Haller, CCC Chief Data Scientist, VP and Co-Director of the Innovation Foundry John has been with CCC for 7 years responsible Data Science and the Innovation Foundry (a partnership between the business and technology. With success in a handful of predictive and AI products in use by top 10 carriers and 35+ total customers, John has proven his approach to research and product management. Prior to CCC, John was at HP doing operations research and general management consulting at McKinsey & Co. John holds an Engineer degree and a MSEE in electrical engineering from Stanford University where he wrote his thesis on the modeling and application of random motion. John also hold a BSE from Princeton University where he graduated with honors and built one of the first self-driving cars in 1997.