Skip to main content

Voicemail Transcription

Posted by mattdarnell on Mon, 04/30/2018

Has anyone tried this with Thirdlane? http://nerdvittles.com/?p=25136
Seems like most of the work has been done, just need to plug it in at the right spots.

Two out of every three prospects ask us about this, the transcription is so much better than it was just five years ago.

-Matt


Submitted by mattdarnell on Tue, 05/01/2018 Permalink

Alex,

That could be a revenue stream for you. If something was baked into the GUI and solid you could get a share of the revenue from the outboard service(s) that you integrate with.

It would be great to activate it on a user by user basis.

We would really be interested in this, I am getting a lot of pressure from our sales folks to come up with a solution.

This is in the same realm as integrating the Connect client with texting, it could be a stand out feature now but it will soon become mandatory.

Submitted by thirdlane on Tue, 05/01/2018 Permalink

Would you as a service provider use your own account for all tenants or allow/expect each tenant to set their own account with the transcription provider? And how would you like to manage customer preferences re transcriptions? Doing this on tenant by tenant and user by user basis (where the end users decide whether they want transcriptions) may be too much - perhaps enabling/disabling this on a tenant level would be a reasonable compromise? That said, we allow setting of "vm by email" option on a user level, so the transcription option does belong on a user level. Your thoughts?

Submitted by matthewmalk248 on Thu, 05/03/2018 Permalink

We are currently paying a third-party a dollar a voicemailbox to do transcription for us, would much rather pay you Alex :)

Submitted by mattdarnell on Thu, 05/03/2018 Permalink

On our MTE system I imagine we would have a single account and charge the customer based on how many users they want. A checkbox in the user voicemail tab would tell Thirdlane to stream the voicemail to the transcription service and then add the result to the email being sent.

STE would set up their own accounts and have some kind of reoccurring billing.

Submitted by eeman on Sun, 06/03/2018 Permalink

Nexiwave, IMO, is the way to go. The other platforms are all-or-nothing. They intervene at the mailcmd= option of voicemail.conf. It will send all the messages to be transcribed. Those mechanisms charge based on time, roughly $1.20/hr of audio. Its estimated that each user will consume about $4/mo in this scenario if they get 15 messages a day, and 22 work days per month. Unfortunately there is no way to bill the customer based on volume, which means that the carriers have to eat the charges.

Nexiwave, on the other hand, is so easy to implement on a user-by-user basis and they bill on a flat mailbox rate. This is more flexible for MTE deployments. One simply puts in the nexiwave email address into thirdlane and then Nexiwave will transcribe the message and email it to you. It even offers the option to convert the message to MP3 format for those using Android devices who has to do a workaround to play gsm wave attachments.

For those who think this could be a revenue stream for thirdlane, you simply don't understand the mechanics of how this service works. There is a reason this is done on IBM Watson. A few companies out there are simply re-branding Google's transcription service via their supercomputer and charging for the service. This is not something you can just throw on a VM somewhere.