I was not sure where to post this, so I ended up posting it here...

I need to know if TeleFlow is able to interact with a digital Nortel Meridian PBX using 2 line side T1s, that is:

Provider's T1 -> Meridian -> IVR

So, what's the recommended telephony hardware for this setup? NMS? Dialogic? Which model in particular?

My customer has a Symposium Express Call Center in place working with the Meridian PBX to manage a small (17 agents) call center. I would then need to eventually transfer incoming calls to that call center through the Meridian... Can I use a simple Transfer Call step to acomplish this?

Finally, my customer also has a Nortel Communications Server 1000S in place, though I'm not sure what it is for (yet). This one supports SIP and H.323. I know TeleFlow does not support H.323, so I guess I would be able to use SIP to transfer calls if they ask me to route through this server instead of the Meridian (I'm just guessing here)...

Do you have any previous experience/testimonial in using TeleFlow with {some NMS|Dialogic digital board} to transfer calls to a Nortel CS 1000S via SIP?

Thanks in advance.

PS: I just received confirmation that my customer does not have a Nortel Communications Server 1000S in place, so forget about that question.

Considering the fact that this project is in the Dominican Republic, what parameters, specs or signaling should be suggested/asked to the T1's provider?

Again, thanks in advance.

The NMS CG 6060 card is the one you will want for this application. This model will handle your current needs: "CG 6060/11-2L/2TE media processing board (2 RJ-48Cs)". It has two T1/E1 interfaces (RJ-48C), and also two network interfaces if you wanted to use it for SIP.

Considering that you want to connect to a Nortel Meridian and that you want to be able to transfer calls through it, you want the NI2 variant PRI, as per this document: T1_ISDN_Service_Specs_NMS

You ALSO want to specify that you want the 2B-Transfer service enabled on the circuit from the PBX to the IVR. This will allow you to use the Transfer Step to transfer calls via the PBX. If you cannot get 2B-Transfer, you can use the Switch Step to connect two TeleFlow ports together.

Ok, just to confirm that I understand this properly:

- If I get the 2B-Transfer service enabled I would need a single T1 connection from the PBX to the IVR and when I want to transfer a call I would use a Flash Hook + Transfer steps, effectively freeing the port

- If I don't get the 2B-Transfer service I would then need the two T1 connections to the PBX (for inbound/outbound calls) and when I want to transfer a call I would use Place Call (on some outbound port) + Switch Steps, silently monitoring both ports to determine when one of them hangs up

Right?

That is essentially correct, yes.

Except the Transfer step does not work like a transfer on analog phone circuits. No Flash Hook, for a start. But the big change is that you need a 2nd port to use Place Call to get your other party on the line. Then, with both parties on line, use use a Transfer step on one of the ports to instruct the PBX to connect the two bearer channels. This is where we get the name 2-B Transfer.

We use the "Transfer Demo 2" example applications (In TeleFlow's "New Application..." wizard, under the "Example" tab) and change the Switch step to a Transfer step.

When the two parties are connected together, the IVR will be dropped from the call, free the ports to take other calls.

Your understanding of the 2nd method, using the Switch step is correct, and is the functionality demonstrated in the "Transfer Demo 2" example.

So, assuming I use the full 24 incoming channels (well, 23 in fact) in the first scenario for the IVR application:

- If I get the 2B-Transfer service enabled I would also need the two T1 connections from the PBX to the IVR and when I want to transfer a call I would use the Place Call on some port of the second (outbound) T1 to get an agent online and then use the Transfer Step on the inbound port, effectively freeing both ports while the PBX keeps the customer and the agent online.

However, if I use just 18 channels for incoming calls leaving 3 for outbound/transfering of calls and the other 2 for outbound faxing, I would then need just one T1 connection from the PBX to the IVR and apply the same logic described above (Place Call + Transfer steps)...

Right?

I need to clarify this because our customer needs to perform an upgrade on their PBX and I need to be clear on how many T1 boards they have to set up on the PBX (one or two) depending on their need to attend up to 18 or up to 23 simultaneous calls... the use of one or two T1 connections between the PBX and the IVR translates to a cost issue that they need to consider.

This is correct. Either of the configurations you described will work.

If you have just one T1 from the PBX to the IVR, then the T1 will have to be configured to allow both in and out bound calls. Also, the hunt order is important. It should be top-down, least idle. In other words, in-bound traffic should use port 1 if it is free, and port 2 if not, etc. This will fill up all 18 in-bound ports before you have a call offered to one of your out-bound ports (which won't be at the Wait For Call step, and won't answer that call).