Available integration for an outside IVR

[cranius28]

Hi all,
This is my first post, and I am somewhat new to the Aspect world.

Here's the deal: I have been tasked to replace an existing DTMF IVR developed in the Aspect CSS/eArchitect, with a speech recognition IVR written using Microsoft .NET (completely outisde of the Aspect development environment). My IVR will reside on a separate server, connected "behind the switch" via lineside T1 interfaces.

My question is, if my IVR has to transfer the caller to an agent, my only option is to perform a blind-transfer back to the Aspect ACD. Knowing this, are there any existing (Aspect) API calls that I can use to pass the information that I gathered in the speechrec IVR and use it to populate screen-pop information for the Aspect to use? I would like to reuse the existing routing/screen-pop routines on the Aspect, using the data that I collected outside of the Aspect environment. If there's no API calls, is there access to a native database where I can attach information (like an account number) to the call record (again for use by the Aspect screen-pop routines)? Kind of like you would do use Intel NetMerge's CIM database to attach information to each unique call record.

Finally, in the absence of any application/database connectivity, does anyone know if I supply extra digits in my dial string during the transfer (these extra digits would be the account number), will those digits be automatically truncated (lost) in the transfer, or is it possible for these extra DTMF tones to be collected somehow and then used for screen pops using a modfied Aspect eFlow?

Thanks for any/all comments.
Mark

 

[AppSpecialist]

Hi Mark... so we are not the only ones.

We are currently investigating doing the same thing, go with Microsoft Speech Server or implement Aspect's CSS. Right now we use the IVR capabilities of the Aspect ACD.

I am interested to know why you are going the Microsoft route as oppossed to the Aspect route... maybe cost ?

Anyhow, we investigated a fair bit as to whether another customer out there has MSS and Aspect, but found none. We figure we would have to end up getting external help to design the integration as right now we do not have the resources to do it ourselves.

But there are a number of APIs/custom controls available from different Aspect SDK's that can probably help, but we have not yet investigated further.

I too am very interested to see if you get any responses.

 

[HardCoal99]

You will want to use Aspect's CTI messaging capabilities to transfer information between Telephony devices (your custom IVR $ the ACD). This is assuming you have Aspect's Contact Server.

We use Aspect's IVR and have been able to develop very good SR applications on it. I'm not really sure why you'd want to develop your own IVR as there are many very good comercial IVR platforms available.

 

[cranius28]

The way I understand it, it's about not having 'all the eggs in 1 basket', and the fact that the IVR is only a small part of the overall enterprise software (in which the rest is written in .NET). It also doesn't hurt that Microsoft has a very good speech rec server that is priced aggressively too.

Cranius28

 

[chris13]

if you could pass a single key-value during the transfer from the IVR to the ACD, you could use the contact-server to do the rest. Passing the key might be done by customising the ANI/CLI during transfer. We do this but use E1/Q.sig to connect ACD+IVR.

IVR:
- genrate key-value
- store results of IVR-dialogue together with the key
- transfer + pass key to ACD

CS:
- use key to query IVR-DB (or whatever storage) for the rest of dialogue results
- assign values to call-variables or do something else

Don't ask me for details on contact-server, it's just an idea...

 

[genki]

first consider where your calls come from and where they need to go. if your IVR is behind the switch you mean the Aspect ACD ? If all calls go to the ACD first and are transferred to the IVR, the new IVR also needs to provide trunks (T1 Channels or whatever will physically endpoint). This endpoint of the .Net IVR needs to receive and send voice data. I don't know the MS Speach Server, but I don't believe MS provides you with the necessary hardware. Typically Intel Dialogic T1 cards are common for this purpose.

The transfer to agent is easy if the call came through the ACD. A simple TCR message (Transfer Call Request) through the Contact Server API (C++, ActiveX, Java) transfers the call to a CCT. With this TCR you can pass 5 variables (A-E) for screen pop purposes.

Other type transfers are more complicated.

 

[axspect]

Your world shouldn't have to change that much. All you are adding is speech recognition to your existing IVR capability. The data passed back to the Aspect for a screen pop will be the same, so just get a good developer and don't let him/her take too many shortcuts.

Can VOIP also help in place of TIE lines?

Third choice, put the new IVR in the cloud and then coordinate the transferred voice call to the Aspect with a data link to the IVR server.