Collin Davis, CTO of Pindrop – Interview Series

Collin is the CTO of Pindrop. He has a background in both Voice and Security. Collin spent the last 8 years at AWS where he ran multiple cloud services at a large scale. During his time there, he founded Alexa for Business and took it from conception to launch in order to help customers voice-enable their conference rooms, lobbies, hospital rooms, hotel rooms, and even warehouses with Alexa. Collin strongly believes in the potential of voice computing and that we are just getting started. After graduating from UCLA with a degree in Computer Science, Collin started his career as a Software Engineer at Symantec. He eventually became the VP of Engineering for the Norton consumer security product line, where he worked with his team to come up with new ways to catch bad guys and protect customers against threats.

What initially attracted you to wanting to work with voice recognition and voice assistants?

I enjoyed computers and programming from a very young age, so I went on to study Computer Science at UCLA. While at school I had the good fortune of getting an internship at Symantec, where I worked for more than 12 years. From there, I joined Amazon Web Services and launched the Amazon Echo. That’s when I first got excited about the potential of voice, what it was doing for smart homes and how it could transform businesses, so I went on to found Alexa for Business to help create smart office spaces.

Years later, I was thrilled when I met Pindrop’s CEO, Vijay Balasubramaniyan, as Pindrop gave me the perfect opportunity to combine my security background from Symantec and my passion for voice technology.

As the founding GM of Alexa for Business, you led the service from conception to release. What were some of the key lessons that you learned from this experience?

Working on Alexa for Business taught me the importance of identity. With any new technology, it becomes more useful when there is a strong way of identifying and authenticating people. Phones, for example, became an even bigger part of our lives and work when we were able to make transactions through mobile applications tied to our identities and passwords. This opportunity that passwords provide for authentication is the same opportunity voice provides for a variety of new technologies and experiences.

Your current role is Chief Technology Officer of Pindrop. Could you briefly explain what Pindrop is?

Pindrop is the global technology leader in voice security with a singular mission to provide real-time authentication and security on every voice communication and voice device. Pindrop’s patented technology and expertise extracts intelligence from calls and voices encountered in the contact center, and across a wide range of devices and applications, to detect and prevent fraud as well as authenticate customers.

The company’s proprietary Phoneprinting and DeepVoice technologies are trusted by eight of the largest banks, five of the largest insurance companies, and three of the top brokerages in North America.

What is social engineering in the context of information security?

Social engineering is a tactic fraudsters use to manipulate people into transferring their funds or use their personal information to take over their victim’s bank account. It’s often used in a fraudster’s bag of tricks to help circumvent call center agent procedures, such as knowledge-based authentication (KBA) methods. This personal information obtained, like maiden names, social security numbers, and more can be easily uncovered to bypass those KBAs.

How is machine learning used to detect social engineering attacks?

Pindrop’s technology monitors all inbound calls and leverages real-time risk analysis to help alert call centers of any at-risk customer accounts that show signs of account takeover.

The technology uses machine learning models to analyze more than 250 biological and behavioral voice characteristics, like the harmonics and frequency of speech as well as rhythm, tone, and emotion. The technology helps identify anomalies in voices to strengthen call center defenses against fraud by detecting suspicious call patterns and account takeover attempts.

Pindrop has reduced fraudulent calls to call center agents by 80% and has prevented $1.2 billion in fraud loss since 2012.

Could you explain Phoneprinting technology?

Pindrop’s Phoneprinting analyzes more than 1,300 audio features to create a complete profile of a user. It combines the full audio of a phone call, including signal noise, codec, and more, with device type, carrier, and geographical information to help call centers identify malicious actors and authenticate legitimate callers.

Call centers are able to use Phoneprinting to identify odd behavior from first-time callers and identify repeat offenders. The technology deters social engineering and can detect multiple callers that are associated with the same phoneprint to identify and track fraud rings.

Why is this technology an essential element of cybersecurity?

When financial institutions and insurance companies receive calls to their contact centers, they need absolute certainty that the call is coming from their customer and not a fraudster.

Pindrop’s technology enables automatic voice and call authentication services, helping to detect fraudulent calls in the background while enabling a seamless customer experience.

Are there any personal anecdotes that you can share that influence your approach towards voice technology and security?

One day, my two-year-old son said, “Alexa, set a timer for two minutes,” when he was put in time-out. To me, this was a sign of a “voice native” generation. Now, whenever he comes across a non-voice-enabled device, he assumes it’s broken. My son’s interactions with Alexa are an indication that voice is a core part of our identity, and that voice technology that can identify a user is game-changing for security and authentication.

Is there anything else that you would like to share about Pindrop?

In an innovative and industry-first move, this year Pindrop extended the use of its voice-based security technology beyond contact centers to consumers on over the top (OTT) streaming devices, services, and platforms. This is the first truly commercial application of the technology, opening-up avenues for its use across IoT applications whether in automotives, voice assistants, or others.

Via its recent partnership with Pindrop, TiVo now has the ability to serve customers with a personalized viewing experience that recognizes and recommends content based on an individual viewer’s unique voice. By leveraging Pindrop’s voice biometrics and voice authentication solutions, media streaming and smart TV devices and services can enable advanced personalization leading to the best user experience possible. This combination of voice and entertainment increases consumer engagement via highly personalized content, resulting in higher retention and less churn.

Thank you for the great interview, readers who wish to learn more should visit Pindrop.