Key Capabilities / Features

Pindrop’s Deep Voice® Engine combines machine learning and AI algorithms to remove background noise and enhance audio to focus solely on the user’s voice. In addition, Pindrop’s liveness detection capabilities and anti-spoofing countermeasures help identify deepfake attacks ranging from text-to-speech, voicebot, voice modulation, and voice conversion, as well as recorded voice playback attacks. In addition to this, its award-winning Phoneprinting® technology analyzes over 1,300 acoustic features of a call to detect any subtle anomalies that help to recognize customers, identify fraud, and determine its device type, geo-location, and carrier.

To further manage account risk, Pindrop also utilizes Trace® technology which leverages the power of graph analysis to identify connections in the Interactive Voice Response (IVR) across time. It highlights which accounts are most at risk, and finds patterns as far as two months before a takeover is attempted. This provides protection for a customer access point that is often overlooked and underestimated as a target for fraudsters. Through a set of APIs, the Pindrop solution can help to identify the most at-risk accounts up to 60 days in advance of an attempted account takeover.

Launched in 2024, Pindrop Pulse Inspect is the latest addition to Pindrop’s cutting-edge deepfake detection Pindrop Pulse product family and can detect AI-generated speech in any digital audio file with 99% accuracy. Pindrop Pulse Inspect extends Pindrop’s technology beyond just call centers but now offers a new use case for media organizations, nonprofits, government agencies, and social media networks. Pulse Inspect quickly analyzes and verifies whether media files contain synthetic speech so customers can proactively prevent the spread of misinformation and maintain their brand credibility. Its Pulse Inspect tool has already proven its effectiveness by tracing the origins of high-profile deepfake incidents involving President Biden and Vice President Kamala Harris earlier this year.

How we are different

- Multi-factor: Other solutions rely on outdated identity databases or single factors such as voice biometrics. Pindrop leverages multiple factors (voice, device, behavior, carrier signaling, liveness detection) which provides a stronger layer of security.
- Big picture: Other solutions look at single fraud events or single point of contact. Pindrop uses a network-based approach to concurrently assess relationships between multiple callers and multiple accounts at an enterprise level. This helps uncover more fraud and up to 60 days in advance of the attempted fraud. Additionally, Pindrop has the largest consortium of fraudster data in the US, derived from more than 5 billion calls monitored, and is available for use by all Pindrop customers to detect “any” known US fraudster. Pindrop is also the only US tech vendor that can offer voice solutions in 50 states – with a court ruling that it has FI status under GLBA and is exempt from IL BIPA.
- Deepfake detection: Pindrop has been preparing for deepfakes for the last 8 years. Pindrop can detect multiple types of deepfake attacks, is the top performer in multiple ASVSpoof challenges, and has 250 voice security patents including 11 patents on deepfake detection.