Verify

AI solutions for imposter-fraud risk detection in a frictionless, cost effective, and privacy preserving approach

As account takeover, Identity theft and fraud proliferate in critical sectors, existing solutions are not keeping pace with the tools, tactics, and traits of bad actors. VoxEQ reduces Account Takeover (ATO) and fraud incidents, false positive rates, and agent time per call for Banking, Financial Services and Insurance Customer Care Call Centers, and the Business Process Outsourcers that serve them.

Key Features

No Enrollment

Works day one by supporting identity verification without enrolled voiceprints. Know about your caller on the very first interaction

Differentiated AI

Among other things, our advanced AI models deploy a high dimensional vector to hone in on a diverse set of traits that reveal identifiable characteristics about the voice. That, and other capabilities, make our system hard to deceive via impersonation, human or synthetic

Privacy Preserving

No data sharing required for VoxEQ to deliver reliable, relevant and timely information about the caller/speaker

Real Time

Works with only a few seconds of voice. Five seconds of speech with insights before the sixth

Cloud Native

Full suite of tools available on latest cloud integrations with leading cloud providers

No Data Storage Required

No required voiceprint or caller insight storage needed. No risk of compromise from data breach or attacks

Regulatory Compliant

Serving banks and financial institutions. VoxEQ works without any data transfer or PII issues

Easy Setup

Connect to an audio stream and get an immediate feed of caller insights from Verify

Proprietary Models

Our AI models are custom built to analyze voice. Unlike others, we don't license or use publicly available AI models that can be spoofed. This means we are better able to stay ahead of bad actors

Language Agnostic

Because we are listening for the anatomical and physiological traits revealed through voice, our products work in any language and can be deployed anywhere in the world

Review Archive of Past Calls

Sanity check that enrolled voices linked to accounts are not poisoned or assess quality of measures escalated suspected fraudsters

Multi Factor

We deliver insights on a broad set of characteristics, above and beyond passwords and voiceprints, to defend against fraud - including gender, age and much more

How Verify Works

An inbound call is answered by the agent or IVR

Via its API, VoxEQ analyzes the callers voice in the first seconds of the call

In real time, VoxEQ's Verify provides multi-factor signals to inform fraud risk elevation

Agent continues with caller confident that the voice is likely the account holder or moves the call through the fraud escalation process

Read our Verify Case Study Genesys AppFoundry Description

VoxEQ's Verify offers the markets most accurate fraud prevention tools and defends against synthetics and other deep fakes

Example

Since integrating Verify, we've seen a significant reduction in verification times and a marked improvement in accuracy. The automated system is incredibly user-friendly and seamlessly integrates with our existing platform

- Customer

We spend a lot of time talking to, and working with, large banks, insurers, online platforms, and others. They all have critical fraud and security issues and though there are products in market to help, many find that these products are not up to the task; too many false positives, too many false negatives, too opaque, too rigid in tuning for their specific business needs, or too large for the level of service or customization required. VoxEQ is different. Reach out for a demo, to discuss a pilot engagement or to learn more. - Jack Caven, CEO

Most of what people refer to as AI, is not AI. This s AI. - Genesys Financial Services Executive

FAQs

A few of our most common question but please reach out for more detail.

How is Verify different than Pindrop or Nuance?

Verify differs in a key ways from competitive offerings. We focus on what the voice reveals about physiology and anatomy to see more of the person behind the voice - something speech to text, pitch, tone, or device traits cant see. When used with our industry leading voiceprint, Verify offers multi-factor benefits -in real time- that others cant, to provide a more accurate and actionable insights.
Outside of the technology, unlike others, Verify does not require any pre-enrollment or knowledge of the speaker and our tools can be customized to meet a clients unique cost-benefit/risk tolerances. Verify allows customers to dial up, or down to cast a wider or narrower net on identifying bad actors. In addition, Verify does not need to store data and we do not need any personally identifiable data (PII) to give clients the most powerful fraud tools available.
Some others offering fraud prevention tools are expensive and so we use them sparingly. How do you compare on cost?
We’ve tried other solutions and found that we get too many false positive to handle without a mechanism to make adjustments How does Verify address this?
How much time does Verify need to assess the fraud risk of the call?

For one party calls, Verify needs only a few seconds of voice to deliver its analysis. After four seconds of audio, Verify will deliver its results by the fifth second. Our cloud based, API service is real time and affords agents a fraud risk insight that is fast and friction free so your customers are not inconvenienced with multiple knowledge-based question and your agents can proceed with confidence that they are speaking with the proper account holder.
What do I get from Verify and how might I take action?

Verify is a cloud-based solution available via real-time API or user interface with different display options to meet your specific needs. We provide caller demographics including Gender, Age, Height and other features. These insights are presented to agents before they engage with a caller and support quick 'friction-less' determination of whether fraud risk is elevated and an escalation is required or whether an agent can reliable work with the caller to deliver the best customer experience
We are moving to cloud based services. What platforms are you on?
Can you help defend against synthetic or deep fake attackers ?

Yes. Our recent testing against deep fakes, including well-publicized examples (e.g. President Biden, Scarlett Johanssen, Andrew Ng's Linked In challenge), were successful in determining that the voice was a fake. That said, we know that as technologies improve, catching bad actors and fakes will be an 'arms race' that VoxEQ is best positioned to stay ahead of.