I am recognized by my voice
GIF uses Voice Biometrics, a technology that guarantees the most advanced method of authenticating people. It only takes a few seconds of the person’s voice for the system to analyze, with 98% effectiveness, who is the sender of the audio.
Each individual, no matter how many times they speak the same phrase, will present intrinsic characteristics that makes each individual unique. The voice is like our Voice print. It is enough to identify how the words are pronounced, the speed and the accent to know who is the person behind the voice. In addition to the aforementioned, the shape of the mouth and breathing are also included among other 50 determining factors.
The great advantage for our clients is the fact that, with Voice Biometrics, the authentication process time is reduced 4 times. On top of that, we have real-time authentication, which promotes much more security for clients and end users, in addition to validating the contact with an agent without having to repeat phrases over and over again. It is also possible to accelerate the service time by an average of 40 seconds per call, ensuring efforts for income-generating activities.
By using the Voice Biometrics services, it is even possible to passively register customers in the system during the call and create an anti-fraud database, which will be very useful for the security of commercial activities and for interactions between the end user and the brand.
Frenquently Asked Questions
Voice Biometric can detect a person's identity by the way they speak. Several attributes contribute to this identification, such as: pronunciation, accent, the way the mouth moves, breathing, etc. Both the physical and behavioral components are evaluated. These functions are known as "voice Print". The easiest way to understand it is to make an analogy with digital print. The way we speak is also unique, and GIF uses technology that can identify anyone.
Voice Biometric can be performed in two ways: dependent text and independent text.
Some people confuse voice authentication with voice recognition. The main difference is that voice recognition associates speech with text for later use in a context that is not related with identifying who that person is. Voice authentication, in the other hand, has the clear objective of using technology to know exactly who the speaker is.
The first step is to register each person on the platform. This procedure is done by capturing the person's voice, speaking the same sentence multiple times. In real time, the system analyzes if the sample has the qualities, we need to use the data later. If not, we ask the person to pronounce again the sentence until it is possible to identify that it is suitable for our clients.
After registering, in the voice capture process, our system is able to recognize who the person is in seconds. This is because in real time it analyzes the database and compares the two sounds.
We use two types of voice biometrics: active online and passive offline. The client indicates which is the most suitable for their needs.
Active - Online: It performs voice authentication in real time, during the call with the person, which can be done actively, online and dependent text.
Passive - Offline: It is used when it is necessary to verify the call recordings and therefore, identify and / or prevent fraud.
GIF aligns the experience of professionals who know the most advanced investigative techniques based on the technology to identify and prevent fraud. Offline passive voice biometrics is used through powerful servers, which record and store digital and contactless sales audios.
Our clients receive cluster reports, in addition to creating blacklists and metadata that directly contribute to the identification of fraudsters. With this information we are able to produce efficient cybersecurity strategies and always in line with technological advances.