Real-Time Speech-to-Text and Sentiment Analysis Using Watson Services
Explore how Nodap streams audio to Watson for real-time transcription and analysis, enabling instant code execution based on speech content.
File
How to transcribe and analyse a phone call in real time
Added on 09/02/2024
Speakers
add Add new speaker

Speaker 1: You have a trial account. You can remove this message at any time by upgrading to a full account. Press any key to execute your code.

Speaker 2: Please dial the phone number that you want to call, followed by pound.

Speaker 3: Hello.

Speaker 2: This is a bit weird.

Speaker 3: Yes, this is a bit weird. Let me explain what's happening. The audio for each of our ends of this phone call are being streamed separately to a small Nodap.

Speaker 2: OK.

Speaker 3: That Nodap is streaming those audio streams to the Watson speech-to-text service. Got it. And then it's getting sent back transcriptions of what we're saying.

Speaker 2: Makes sense.

Speaker 3: The Nodap is then using a simple web socket so that it can display the transcriptions of what we're saying in this web app.

Speaker 2: OK.

Speaker 3: As you can see, this is all happening more or less in real time. The speech-to-text service is configurable, though, so you can choose if speed or accuracy is more important to you. So for some people, maybe getting the transcriptions as quickly as possible is going to be a priority for them. Or maybe other people won't mind a second or so delay if it means that they get a high level of accuracy.

Speaker 2: Which is this?

Speaker 3: We're going for speed here, so some of these transcriptions aren't exactly right. But just displaying transcriptions probably isn't the most useful thing we can do. Once we've got the text for what we're saying, we can analyse it, and then we can trigger code automatically based on that analysis while we're still talking.

Speaker 2: What sort of analysis?

Speaker 3: Well, by sending that transcribed text to the Watson Natural Language Understanding Service, we can do all sorts of things. Like entity extraction. We can recognise when we mention certain things, and then we can trigger code if one of us mentions some certain type of thing. Or we can do things like sentiment analysis or tone analysis, analysing the words that we choose. And all of that is what's being displayed on the right-hand side here. Call me again, and let's give that a try and see if we can get the sentiment analysis to recognise a bit of a change. You have a trial account. You can remove this message at any time by upgrading to a full account.

Speaker 1: Press any key to execute your code. Please dial the phone number that you want to call, followed by pound. Please dial the phone number that you want to call, followed by pound. Please dial the phone number that you want to call,

Speaker 2: followed by pound. Please dial the phone number that you want to call, followed by pound.

Speaker 3: Hello. You are a terrible person. I hate you. I'm fed up with everything that you do.

Speaker 2: You're unkind.

Speaker 3: You're horrid. You smell. I've had enough of your rubbish. That's me. You're so annoying. You're so irritating.

Speaker 2: I didn't see this coming.

Speaker 3: Why can't you just stop it? I can't put up with this rubbish anymore. That's probably enough of that.

ai AI Insights
Summary

Generate a brief summary highlighting the main points of the transcript.

Generate
Title

Generate a concise and relevant title for the transcript based on the main themes and content discussed.

Generate
Keywords

Identify and highlight the key words or phrases most relevant to the content of the transcript.

Generate
Enter your query
Sentiments

Analyze the emotional tone of the transcript to determine whether the sentiment is positive, negative, or neutral.

Generate
Quizzes

Create interactive quizzes based on the content of the transcript to test comprehension or engage users.

Generate
{{ secondsToHumanTime(time) }}
Back
Forward
{{ Math.round(speed * 100) / 100 }}x
{{ secondsToHumanTime(duration) }}
close
New speaker
Add speaker
close
Edit speaker
Save changes
close
Share Transcript