We use NLP (Natural Language Processing) to convert speech to text.
One of the services we provide within the platform is real time speech to text. While this is pretty straight forward - you’ve probably done this on your phone for emails or texts before - there are a couple pieces of background we think you’ll find useful.
- Just like when you try to use it for a text on your phone, our transcription services can get things wrong.
- In the next lesson, we’ll show why this really isn’t a big deal, but it’s worth pausing to think about WHY the transcriptions are wrong.
-
- Fast talker
- Different dialects
- Blended words
-
For example, Allergy vs algea; telemedicine vs tell him medicine. When you look at it, it’s understandable!
- Our guidance here is first and foremost to use it as an opportunity to build your employee’s self awareness. When we test learners who speak slowly and articulately, they get perfect or near perfect results. So you will not be alone if you simply tell your learners to ‘slow down and articulate.’
- Our next piece of guidance is that you should embrace the alternatives that transcriptions can provide- which leads us into the next lesson and the Importance of Coaching your AI