Avaya was born out of AT&T/Lucent's legacy, but since its 2000 launch, the enterprise telephony vendor has tried to recast itself as an enterprise applications company with a focus on voice. Recent moves include the migration of Avaya's legacy PBX to a Linux-based server application, and the introduction of an application server for partners and users to develop VOIP-integrated software such as applications that integrate VOIP and messaging with ERP software, Web sites or portals. We recently discussed the company's evolution, as well as current trends in the enterprise VOIP market, with Avaya CEO Donald Peterson.

How does Avaya's IT and applications approach to telephony affect Avaya's partners, customers and its own product development?

They're either going to change their businesses or struggle with the evolution. That's not a new story. There is still a TDM installed base that needs to be serviced, and there will be a significant revenue opportunity for folks. But the longer-term value opportunity will be in this new world. By creating our own application server, in some ways we're trying to give an advantage to people who have written code for Avaya telephony products in the past. In the future, we will put the call manager and communications manager into an [IBM] WebSphere-like platform, where people would write their applications. That will most certainly happen over time.

As Avaya has moved deeper into software and applications, Microsoft has become more interested in voice technology. Does this worry you?

That's absolutely on our radar screen. Microsoft's Live Communication Server does call control, so in that sense it's a direct competitor. But call control long ceased to be the only offer in a PBX. There are 800 features that were created at different times for different customers that we support. Microsoft will be awhile catching up with that. Having said that, there are a number of users, particularly at the small end, where the LCS [Live Communication Server] could catch on. Certainly in the beginning, it's cheap dial tone. Microsoft has a huge network of developers that you would expect to be writing against that. It's going to be a rapidly developing area. But telephony is a non-trivial exercise. So I don't think I'm worried about large companies turning to LCS in a wholesale way in the near term.

Is the integration of voice over Wi-Fi and cell phone technology going to happen soon?

Keep in mind the cellular network is a TDM network. There's certainly some [technologies] that we're working on. We've announced partnerships with Proxim and Motorola, where a dual-mode handset hooks up to a hot spot. Our recent Nokia announcement goes the other way; it leaves the phone on the cellular network, but it extends the PBX call control out to treat the cellular phone as an extension to the PBX. This retains applications that are on the PBX, like call recording and those kinds of things. There will be other evolutions like that. Right now they're not dependent on an IP transport out to the phone. But that kind of thing would enable streaming video, videoconference calls on the phone more readily than the TDM does today.

With so much focus on software and IP telephony, what is the state of TDM technology support and product development?

We treat TDM as an investment to be leveraged. We want to add value to the pure IP world, but we recognise that's not an environment that most of our customers will be going to for a long, long time. If we create a solution that is deliverable in the IP world, we immediately think of how it can be delivered as much as possible by a TDM user on an IP extended link. Our customers can then get a lot of value without having to go through an expensive rip and replacement. They can do that when they're ready, when it makes sense.

There is a lot of interest as to what one particular Avaya customer - Merrill Lynch - is doing with VOIP: deploying a mix of Avaya IP and TDM technology in large corporate offices, and using Cisco VOIP equipment in branch offices. Is this the state that other large customers are in, regarding VOIP and telephony - using multi-vendor IP and TDM technology in separate areas?

Most companies are looking for a more uniform solution than the one Merrill Lynch has opted for. As applications extend outward, that uniformity will become somewhat more valuable. If you make a decision on a cost optimisation side, you could end up with different solutions in different environments. The example you cite includes two extremely different environments - large central locations vs. branch offices with handfuls of people.

Overall, Cisco's approach to VOIP isn't as constructive, I think, as ours. They are on one end of the spectrum, doing very proprietary implementations of quote/unquote, open standards. I would characterise us as on the other end; we're trying to be open so these things can work together.In the long run we'll see what's a better strategy.