Interview: Sees Machine Learning Throughout the Entire Customer Lifecycle

The world is in constant motion in the business realm and so is data and thusly the need for accurate, enlightening and on-going analysis. understands the complexities of this notion and arrives at a solution using machine learning to give their enterprise customers a distinct advantage. We caught up with Jeff Erhardt, CEO of, to learn much more.

insideBIGDATA: The story of is quite interesting. Can you tell us how it got started?

Jeff Erhardt: The company was founded by a team led by Joshua Bloom, a Professor of Astrophysics at UC Berkeley who saw an opportunity to apply technology as a means to better predict the movement of astral bodies in space. Bloom and his cofounders’ interest in using machine learning for “real world” data led to an interest in applying machine learning to business problems. In particular, they saw an opportunity to bridge the growing gap in the implementation and use of cutting edge analytics technologies, particularly for customer-facing employees using other cloud-based tools such as Hubspot, Salesforce, and Zendesk.

insideBIGDATA: Jeff, machine learning is such a fascinating and burgeoning field. What will, as a new company in this market, be all about?

Jeff Erhardt: We deliver machine learning driven customer experience solutions integrated directly with the most common cloud-based business applications. Our applications span the entire customer lifecycle, from acquisition, through monetization, and to retention, allowing our customers the easy implementation of a packaged solution while still maintaining the power to cross their data “silos.” Because the algorithm is based on the latest machine-learning technology it is able to get smarter, and make better predictions, the longer it runs.

insideBIGDATA: How does this set your company apart from those in the same space?

Jeff Erhardt: Our main differentiator from other machine-learning companies is that we’re focused not just on high-performance algorithms, but on delivering an end-to-end application for business users. While we continue to push boundaries of cutting-edge machine learning technology, we made an early decision not to get sucked into the “algorithms arms race.” We hold a fundamental belief that the best analytics technologies will fail unless they can be implemented in a timeframe relevant to the business and interpreted by the ultimate decision makers.

insideBIGDATA: Why and how is machine learning so important to the enterprise?

Jeff Erhardt: The best way to think of machine learning is “predictive analytics on steroids.” Whereas predictive analytics has historically relied on restrictive, so-called parametric models (such as linear regression & logistic regression), machine learning employs truly data-driven models that can extract more knowledge from your data. As a result, machine learning is far more accurate than static business rules or classical analytics, especially for problems involving large amounts (i.e. Volume) of heterogeneous (i.e. Variety) data.  Moreover, because the models themselves are data-driven, they are not limited by the ability of a human to presuppose which factors which will drive the outcomes. Finally, the world does not stand still, and neither should predictions; Machine Learning based models have the ability to self-improve over time. Ultimately, this means that Machine Learning simply provides better answers, and in a data-driven world, better answers lead directly to competitive advantage. We enable business users to find the proverbial golden needle in the Big Data haystack.

insideBIGDATA: Specifically in the realm of Big Data, what does your company need to do in order to secure a piece of this growing pie?

Jeff Erhardt: We don’t make an explicit distinction in how we think about problems associated with Big Data versus those that are not; we believe that sooner, rather than later, the world will return to viewing it as “data.” In particular, we believe that there are tremendous opportunities to be seized by making better production decisions with whatever data an enterprise may already have. With that said, our core technology is ideally suited to address the challenges associated with the Volume, Variety, and Velocity of today’s Big Data. What distinguishes us in this market is our ability to deliver rapid time to value by providing easy to implement customer experience solutions directly integrated with the most common SaaS business applications.

insideBIGDATA: Because machine learning is relatively new, your imagination must be going wild with future possibilities. What are these possibilities and how will be involved?

Jeff Erhardt: Absolutely. On the one hand, great machine learning is already seamlessly integrated into certain aspects of our lives, such as adaptive spam filtering in Gmail or highly customized product recommendations from Amazon. On the other hand, we have barely yet seen a glimpse of power of being able to hear our customers before they speak. Imagine intervening in a support problem before the customer even knows he will be upset, developing the next killer feature without it ever being requested, improving energy efficiency by knowing when, where, and how a motorist will drive, or improving lives by providing cost-effective credit to those excluded from the existing system. We are powering exactly that predictive future.

Resource Links: