What can SCAH do with superhuman* voice understanding?

Michael Devlin
SIDE Labs
Published in
2 min readMar 2, 2022

We recently deployed a Voice AI system that directs SCAH callers to to their local bureau. We are now pondering: what next? What other possibilities does this superhuman voice understanding unleash?

Below are some areas we’ve been considering. We’d love your ideas! Tell us which you think would be useful or, even better, give us other ideas. Any thoughts are hugely appreciated. Please email: michael@sidelabs.org

Frequently Asked Questions

We could offer automated help for certain categories of query. Are there certain categories of questions that could be handled by an automated system? Alternatively, are there screening questions that you’d like a caller to be asked before a human picks up?

Sentiment Analysis

Voice AI is pretty good at sentiment analysis. Could that be useful? For instance: would it help you if you knew a caller was 6/10 on a ‘stressed out scale’?

Out of Hours Experience

Currently, calls get forwarded to local CABs even when they are closed: so SCAH relies on the out of office voicemail at the local CABs. How could this be better? Should we use the Voice AI to provide out of hours information?

Translation
Our superhuman voice system could do real time translation. Would that be useful?

Unsuccessful Calls

Out of the calls to SCAH that we fail to redirect to a local bureau, the majority are silent callers. How common of a problem is it for advisors to deal with silent calls? Does anyone have any tips on how to get silent callers talking?

Better support for local CABs

Sometimes we worry that the Voice AI we’ve implemented at SCAH pushes the bottleneck down the line, increasing the pressure on local CAB. How could superhuman voice understanding help the CABs?

Any thoughts or suggestions — or if you’d just like to chat! — please email: michael@sidelabs.org

* It really is superhuman! There are times where we can’t figure out what a caller is saying, but the Voice AI system correctly identifies the postcode. It’s uncanny.

--

--