How a Full-Text ASR System Improves GRC

David Garrod April 30, 2019

Governance, Risk management and Compliance (GRC) covers a wide range of different practices and needs. Speaking in general terms, good GRC practice requires analysis of the behavior of employees, contractors and other individuals for whom a business has responsibility. If this cannot be done, one can be sure that behavior is taking place which violates standards, policies, and even laws.

Most business communications, such as email, social media, and other written communications, are easy to analyze automatically for violations. The one exception is audio communications, such as phone calls. Even if recorded automatically, it is painfully resource-intensive to analyze all the phone calls made and received by a business. Consequently, GRC practitioners will struggle to detect inappropriate behavior in voice conversations.

Automatic Speech Recognition (ASR) systems seem like the obvious solution to this problem. An ASR system processes spoken language into a written output, in the form of a transcription which can be easily analyzed. However, there are key distinctions to understand. The distinctions between phonetic and full-text ASR systems matter, as they do not have the same benefits for GRC.

Our GRC white paper discusses this issue in detail. It outlines how ASR systems recognize language, explains the differences between full-text and phonetics-based ASR systems, and the benefits of each of them for GRC. Click here to access it now.

David Garrod

David Garrod, M.S.E.E., Ph.D., J.D., is the General Counsel of Voci Technologies, Inc., as well as its designated GDPR Data Protection Officer.

Access our ASR API

With up to 1000 hours of audio at no charge