In a recent Cape Privacy Webinar, A Discussion on all-Things Machine Learning, Data Science, and Data Privacy, Cape Privacy CEO Ché Wijesinghe and Priceline CTO Martin Brodbeck talk about Brodbeck’s wide range of experiences managing and analyzing data for top companies and innovators like Pfizer, Shutterstock, Diageo, and a number of private equity firms. The pair also focus on his current work with one of the world’s leading travel and e-commerce brands.
At Priceline, Brodbeck uses data and machine learning to power customer initiatives. In that process, he deals with a host of technical and regulatory challenges associated with managing the sensitive transactional data Priceline uses for training its data model. To protect consumer privacy, great care must be taken to secure the data that is collected during all the various transactions as customers make and pay for their travel arrangements. That data holds tremendous value for the organization, but because of privacy and data security requirements, there’s tension associated with how the data is protected, stored, and used.
“You want to take advantage of this sensitive encrypted data, but you need to do it in a way that ensures privacy and compliance, and to make sure that the information is encrypted,” Brodbeck explains. “But there's a misconception in the marketplace, where the privacy and compliance pieces are about taking a customer request and then deleting their data.”
Brodbeck explains that protecting private data doesn’t necessarily require deletion, as some believe. To the contrary, at Priceline he’s focused on consolidating the company’s data assets to drive a data science strategy that extracts better consumer and market insights from the data, and to support better decision-making within the company. His team is looking at encrypted learning from Cape Privacy to help do the job.
“The next evolution of data science and machine learning is figuring out how to take advantage of all of this rich data that needs to be encrypted, and being able to share it in a way where you can take advantage of it to drive a lot of your products” while maintaining the integrity of the data and protecting consumer privacy, Brodbeck says. “The big focus for us is to create a customer data platform that allows us to track and capture how our consumers are using our products, do that in a responsible and security-first way, but also create new machine learning algorithms to power things such as personalization, recommendations, and pricing.”
One aspect of the business that presents an interesting challenge—and rich opportunity—associated with the company’s use of data is the fact that Priceline is not a standalone enterprise. Priceline is a part of Booking Holdings, which also owns travel fare companies Booking.com, Agoda.com, Kayak.com, Cheapflights, and Momondo, rental car price aggregator Rentalcars.com, and the restaurant reservation service OpenTable. The data that each of these services collects could have tremendous value for Priceline, but without a way to share the data securely, and to train the company’s data models while protecting customer privacy, that value remains out of reach.
Brodbeck says that companies like Priceline want to share sensitive information across brands, but that, traditionally, it has been difficult to do. What has been tried before are mostly manual processes, like data clean rooms, that keep data in one place, but that don’t often deliver value commensurate with the costs and risks associated.
“That's where the power of the Cape Privacy platform really comes in, allowing you to keep data where it exists within your enterprise, putting in lightweight but powerful legal, security, and compliance workflows on top of your existing data,” Brodbeck says. “That preserves data masking and the security of the information, but in a way where the information doesn't become compromised, while unlocking the value of that information to power things such as data science, product development, and hyper-analytics.”
Brodbeck also reflects on his past experiences in pharmaceuticals, consumer products, and financial services and says he can see how Cape Privacy’s encrypted learning platform could be a transformative tool in those industries as well. In particular, he calls out the potential for encrypted learning’s use in clinical discovery and research within the pharmaceutical and biotechnology industries. There, researchers deal with highly sensitive patient information, and that data could lead to new breakthroughs in drug discovery and drug development if available.
“Unlocking that patient information and leveraging a platform like Cape to keep the security and integrity of guidelines such as 21 CFR Part 11 on top of the existing clinical development would be super powerful,” he says, adding “And in the financial technology space, where you're dealing with highly sensitive information for a risk analysis from customer information through trading patterns is another key area where I think this technology is very applicable.”
Martin Brodbeck’s career as an innovative and successful CTO across a variety of industries allows him to share many insights on how encrypted learning can transform data science into a true competitive advantage. That is vital for organizations in markets where a one- or two-point model improvement can mean millions, if not billions of dollars of value. For the full conversation between Ché and Martin, check out, A Discussion on all-Things Machine Learning, Data Science, and Data Privacy.