Contact centers whitepaper
Find out How Automatic Speech Recognition and Call Analytics Empower Contact Center as a Service Providers
Powerful speech-to-text technology can provide accurate transcriptions of phone conversations. This data can be used to extract valuable customer insights, helping to identify the causes of customer dissatisfaction and revealing opportunities to improve compliance, operational efficiency, and agent performance.
Read the case studyThe benefits of implementing voice technology in a call center include:
- Cost savings of over 30%
- Customer-satisfaction-score improvements
- Effective Sales Enablement
Trusted by
Benchmark
Engine Performance Highlights
Vatis Tech conducted this study to evaluate and compare the leading machine speech-to-text (MT) technologies for contact center conversations. The latest benchmarks showed that Vatis Tech's technology achieved a precision of 95%, which is market-leading for noisy human-to-human conversations.
The gap in the chart is a gap in training data, not in engineering. Telephone audio is narrowband — roughly 8 kHz against the 16 kHz of the broadcast-quality speech general-purpose engines are built on — and models meeting it for the first time lose accuracy sharply. On broadcast recordings the same competitors score far closer to us. Contact center audio is what our engine is trained on, so it does not pay that penalty, and that is the difference these bars are measuring.
The study also found that Vatis Tech's automatic transcription engine produces transcription accuracy levels among the highest in the industry, specifically tuned for contact center conversations. The technology is deployed in 5 languages: English, German, French, Spanish, and Romanian.
The aim of the research was to identify the practical business impact of an accurate transcription, for a contact center.
By accurately transcribing contact center conversations, the analytics module is empowered to provide a deeper level of understanding of customer interactions. This allows businesses to extract actionable insights from conversational data, that can make a positive and relevant impact in the business by improving customer service, reducing costs, and boosting efficiency.
Demo
Watch the engine on call audio
Examples
See for yourself
Vatis Tech's speech-to-text engine delivers a ground-breaking accuracy for every speaker, automatically converting letter-written numerals to number-written numerals and identifying a wide range of entities like people and company names, dates, or locations from your audio files. It has over 95% word-level accuracy.
Thank you so much, Mr. Fairbase. Can you please provide your national insurance number?
National insurance number is 369-4456.
Wonderful. Also, Mr. Fairbase, can I please have your date of birth?
My date of birth is 9-26-1995.
We were impressed by Vatis Tech's ability to quickly and accurately transcribe phone calls, train custom models, and remain price competitive. Our partnership with Vatis Tech allows us to quickly convert speech into text, analyze keywords, incorporate sentiment analysis, and ensure script adherence. This enables our corporate customers to gain actionable business insights and customer experience insights from conversational data.
For engineers who read the docs before the marketing page
Read the documentation, try it free, tell us how it goes.