When AT&T purchased NCR, I moved to the new company and became responsible for marketing the Teradata data warehouse in the southeast. Teradata was a big data pio- neer and was the base technology of the largest commercial data warehouses in the world. Business was storing mas- sive amounts of data — retailers, for instance, were storing every transaction they executed. The internet was producing information and clickstream data at incredible rates. From an analytical standpoint the information available and its growth rate were mind blowing. HCOB: And that’s when you took the plunge into entre- preneurship, right? Hale : That’s correct. My passion for analyzing data and solving business problems led me to join a start-up focused on data mining tools and solutions, Neovista. The technologies we called “data mining” are the underpinnings of modern-day machine learning and artificial intelligence. Big data was growing quickly, and we soon saw large companies stepping forward into leadership roles. One of the founders of Teradata called me while I was a VP at AT&T and said that he was starting a new company that was going to go help customers get more value out of those warehouses. Our goal was to lead our customers into the world of machine learning. There were about 20 people at Neovista when I got there, growing to about 35 when I left. I came to realize that the size and nature of the opportunities I was seeing in big data were best addressed by companies with the requisite scale and expertise required to execute on a larger scale. Given that, I accepted a new opportunity at IBM to run what was essentially a skunkworks operation about big data. It was a great little team focused exclusively on leverag- ing business intelligence and advanced analytics world- wide — you might call it a start-up within a very large operation — right up my ally. I decided that, to meet the needs of predictive modeling in the world of machine learning and big data, we would have to use something other than the traditional techniques. I took the opposite tack of most researchers and started working with simplifying the analysis and relying on the massive amount of data available to improve accuracy.” HCOB: That’s where you developed your patent for creating predictive modeling, right? Hale : That’s right. It’s called “Process and Heuristic Statistic for Prospect Selection Through Data Mining” — quite a mouthful, but what it provides is a way to create
predictive models allowing computers to provide insights without human intervention. Traditional predictive tech- nology was based on a complex, iterative, manual process that takes place in off-line platforms. It required expert analysts and usually took a significant amount of time, and the results of that process could be obtuse and hard for an untrained business user to understand. The patent provides a way for computers to create and implement models “on the fly” in an accessible platform without human intervention. HCOB: Why is that important? Hale : To begin with, artificial intelligence must take place at the point of execution, meaning that offline model creation presents significant obstacles. We simply didn’t have time
Richard Hale says he chose Auburn because he envisioned a technical career. He began his studies in Engineering, but, “ I soon realized that business was my passion and I transferred to the College of Business. ”
46
Powered by FlippingBook