Topic of the month covered by WebRTC “activists”
Last week I published a post covering the release of SyntaxNet as open source by Google.
Learn more about SyntaxNet.
Technically WebRTC and SyntaxNet have nothing in common. What they do have in common is that Google again makes it easy to add capabilities, once in the hands of the rich and experts, as a feature into applications. Once this is possible, Natural Language Understanding (NLU) can be added also into WebRTC applications.
For this month the topic is: Google just released its NLU engine SyntaxNet as open source. How can WebRTC applications benefit from it?
Starting with my opinion on this topic.
SyntaxNet is not a complete solution but rather a technology for machines to understand a language and context of text. This is a complicated piece of technology. Having this technology as free to use open source and combining it with other technologies such as speech to text allows WebRTC application to enhance their capabilities and be “smarter”. Examples include: Analysis of calls for improving service, alerting about customer issues, automatic suggestion of possible agent actions and an agentless independent bot. The use of NLU technology in WebRTC extends far more than just B2C/contact center applications.
Same as any other VoIP application.
You will need to first run speech to text and then use SyntaxNet for whatever reason you see fit. It can be for the purpose of feeding a bot service, checking things such as sentiment analysis, etc.
The thing is that it has nothing really to do with WebRTC and everything to do with language communications.
I think NLU could be combined with recordings in use cases where you need to go back to calls in which a specific topic was discussed. Better parsing of a call transcript should allow for better text search results related to that transcript. That’s an interesting use case, but the bigger impact will be when it’s combined with more intelligent bots who can act as the first line of customer support and only refer a customer to a human agent for more complicated or subjective tasks. Interpreting a phrase like “Let me speak to your manager” with software will become the new “dial 0 to speak with an agent.”
Want to joint Topic of the Month as contributor?
Rules of engagement for Topic of the Month: No product/company promotion. Contributors should have a wide market perspective.
If you would like to join this initiative and have your opinion published here in future posts please drop me a note.