What is a transcription speech recognition?
Until recently, Advanced Speech Recognition (ASR) relied on a person’s use of a ‘grammar’ but while this greatly increased accuracy, it had two major downsides; firstly, the system had to be carefully designed, so callers would say things the grammar expects and secondly, the system was unable to recognise unusual words or phrases from the caller.
The new generation of transcription-based ASR resources like Amazon Transcribe and Google Speech To Text take a different approach. They use enormous amounts of material to create a highly effective language model which recognises normal sentences and uncommon words like place names, product names or even scientific terms - and they can also work without the help of grammar. Using this simplified process, audio is passed to the ASR system, so it can return an accurate transcription that even includes punctuation.
Advanced features can also allow the user to tweak the behaviour by adding extra words to the language model, so that the system can learn to understand and transcribe almost anything.
How does transcription work?
Oration has enhanced the process of Transcription by integrating the very latest technologies including Amazon and Google. Oration’s Transcription recogniser understands a caller's natural language and creates an accurate Transcription before the Oration interpreter begins extracting information from the text. Based on that information, Oration can apply an intent to every request.
Mapping what was said by a customer to what was needed (the intent) is critical to the caller experience. Through an ongoing process of supervised learning, Oration simply surfaces the information to the Oration Administrator, so they can map customer responses and create complete coverage for all the caller intents.
What makes Oration different?
With Oration, you can pick and choose the best Transcription engine for your use case with peace of mind that your contact centre is always operating in accordance with industry best practices. While the Oration Interpreter uses sophisticated technology to understand critical words and phrases and ignore irrelevant speech, Oration’s Supervised Learning capability offers a simple way to improve accuracy by allowing contact centre professionals to view caller responses and map them to intents. The real-time hints feature also gives administrators the power to add any unusual words themselves to give the Transcription engine a helping hand when converting customer responses to text. With the benefit of the latest Transcription technology, Oration’s pre-built information extraction can also identify and surface a range of common data types from a customer’s transcribed speech such as a date of birth or customer number.
- Reduce average handling times
- Identify and Verify your callers
- Increase uptake to self-service
- Provide targeted banners
- Facilitate a digital channel shift
- Improve agent and customer engagement
- Support speed to competency
Discover our features
Advanced Speech Recognition (ASR) enables natural spoken language to be interpreted and converted into text, so it can be used to apply caller intents and extract information.
By surfacing caller responses to an administrator, ongoing supervised learning helps the ASR system learn how best to match intents with the language used by callers.
Intelligent Call Routing (ICR) ensures huge volumes of calls received by contact centres are quickly matched to the right outcome every time.
Graphical User Interface Dashboard allows contact centres to build and manage different customer journeys and generate insightful reports on the entire call routing solution.