In today’s world, customer service widely involves the use of alphanumeric characters and a lot of confidential numeric data. Such characters are frequently used in serial numbers, passwords, OTP, tracking numbers, financial data, etc. Most of the sales guys and the customer service staff can totally feel the uneasiness in the voice of customers while spelling out sensitive information. They often get irritated as well while repeatedly spelling out information to ensure that they are heard correctly. We can across such situations in our day-to-day life, this has become a part of our life.
Since such alphanumeric and sensitive characters are not very common in usual conversations, most of the commonly used speech recognition systems are not trained on such datasets so they don’t process such characters in speech. Even the best speech-to-text engines available today are not capable of high accuracy at recognizing alphanumeric characters and maintaining confidentiality by flagging/masking conversations.
ODIO brings the solution!
At ODIO, we have developed a solution to this problem. Our system is capable of recognizing numerical and sensitive data with high accuracy and precision. The platform has been repeatedly beta-tested against multiple datasets in different situations, resulting in up to an accuracy rate of 95%.
One of the most striking features of our platform is the capability of Flagging a call if a customer has shared sensitive data like OTP’s, passwords, credit card numbers, etc, maintaining confidentiality and ensuring customer satisfaction. When the conversation is transcripted with the help of ODIO’s speech-to-text algorithm, such confidential information gets masked. For eg., if the customer has mentioned an OTP, it gets printed as “XXXX” in the transcription.
Our development team at ODIO has been dedicatedly working on such innovative solutions. To discuss this in more detail, please follow our blogs at Odio or contact us over email at support@odioiq.com
Leave a Reply