Back in '96, WIRED Magazine called antitrust and intellectual property attorney Gary Reback 'the only man Bill Gates fears.' And - real or not - those fears would now seem justified. In '97, working in behalf of Netscape, Reback (along with Wilson, Sonsini colleague Susan Creighton, and others) was among the first to clearly articulate the antitrust case against Microsoft. He helped convince the DOJ to litigate; and his influence - stretching into front offices up and down Silicon Valley and beyond - was crucial in garnering support for the case through its successful conclusion, earlier this year.

Now Reback has quit practicing law and founded a high-tech startup - in the process, handing biz/tech journalists one of the year's juciest storyline hooks. Reback's new company, Voxeo (Scotts Valley, CA - 831-439-5130) proposes to solve the "telephony problem" for web-centric developers. They're building a community, providing low-or-no-cost tools, and inventing cooperative strategies to simplify XML-based telephony application development and remote testing; exploiting popular and broadly-supported XML variants such as VoiceXML and WML (plus a CallXML variant of their own) at toplevel. They're engineering middleware that coordinates between TML (Telephony Markup Language - our generic term for beasts like VoiceXML) scripts and the scary, underlying phone stuff; and permits remote integration of telephony to existing e-commerce back-ends. Ultimately, they aim to become profitable by hosting the "telephony parts" of applications on a national network of servers.

In a sense, it's a "global platform play." But with a distinctly open, Internet twist. Broadly-supported telephony markup langauges at the front end neatly abstract away from the underlying middleware and hardware, eliminating fears of lock-in. The development model is simple: compose a one-call app description, post it, and watch the middleware run it on an arbitrary number of lines. The prohibitive cost of buying a development platform is eliminated, as are fears that the app you write on a small development system won't scale neatly to higher line-counts and traffic metrics. Voxeo has designed their middleware so that (in theory) you don't have to worry about that stuff. And of course, integrating with your existing e-commerce infrastructure is close-to-transparent, since it's all happening at the level of XML scripts browser data-models, and arbitrary URL references.

All in all, Voxeo is hoping that this collection of benefits will attract numerous web developers to incorporate telephony in their e-commerce plans. Ultimately, Voxeo is looking to be the place that hosts your finished apps, or, more properly, their telephony components. They're preparing to do this at carrier scales, of course, and shoulder the burden of maintaining high QoS and availability for large clients. But Voxeo is also cognizant that a huge market for enhanced telephony exists at smaller scales - among small and medium-sized companies who, at present, find the cost of premise-based telephony infrastructure (and the coarse scalability models it permits) prohibitive. They envision developing channel partnerships (and welcome calls from folks presently in the telephony sphere) to develop apps, customize them for individual clients, and deploy them against their platform under flexible pricing models. Telephony developers looking to escape the need to sell a hardware platform under every cute IVR app would be well advised to make that call, today.

Too good to be true? Nope. We visited Voxeo on November 28, 2000, and spent several hours chatting with Reback and co-founder (and former Wilson, Sonsini attorney/engineer) Clegg Ivey about the company, the technology, and their hopes for the future. What came out of this meeting was a sense that Voxeo's vision is well on its way to being achieved: that it's the right vision for the market. We were also pleased to discover that Voxeo seems to be, in no way, about grandiose "global platform plays," big egos, or the cult of personality. Reback and his colleagues are open, humble, accessible, competent people - generous with their time, eager to articulate a very sober vision, very serious about building a company and solving problems for an industry. They know - and acknowledge - that what they're doing is hard. And as you'll see from their remarks, below, it's hard not to think they're going about it the right way.

Computer Telephony: How did the idea for Voxeo come about, and when did you decide to start the company?

Gary Reback: Originally, I was going to do a b-to-b exchange. I had done essentially all I wanted to do as a lawyer, and I was tired of showing up at companies all the time and not having anybody happy to see me. So I started asking around to all my friends, including Clegg Ivey, Voxeo Director of Technology Development and co-founder, to find a technologist who knew about XML and NOW Network of Workstations implementations, because that's what exchanges were built on. Clegg told me that he knew Jonathan Taylor CTO of Voxeo, who had done the NOW implementation for Visionix, a sister company of Inktomi, and he offered to introduce us.

After that, Jonathan and I talked on the phone, and I asked him to fly out here so we could meet, because he was in Orlando at the time, on vacation. So he basically put a couple of shirts in a knapsack, and flew across the country the next day, and we had a meeting where I explained to him what I wanted to do, and asked him what he thought of the idea. And he sort of shrugged. So I asked, well what are you interested in? And he began to tell me about all the pain and suffering that he and others he knew had encountered over the past six years as they had tried to extend web capabilities to the phone platform... I was aware, of course, of the great trends sweeping all of technology, and how the web way of doing things, in terms of markup and standards was changing the whole of software development. His point was that this was nowhere more true than in telephony, and he argued that over the next couple years telephony would start to change, a large portion of it going to markup languages and web protocols, instead of the traditional proprietary APIs.

We thought about whether we should try to do the next hot application - the next unified messaging, the next follow-me/find me, the next virtual call center. But we all knew, coming from the web, that everything starts vertically... and then eventually horizontal layers are created, where somebody like Exodus or Verio provides the infrastructure that everybody then builds their vertical businesses on top of. So we said, this is going to happen here too, eventually, so why don't we just leapfrog that vertical stage, and focus on infrastructure? Because we knew that if we could solve these problems for one application, we should be able to solve them for a variety of different kinds of apps. So we thought we should try to become the Exodus of this space, or, more correctly, the Akamai of this space, so that people who build the vertical apps wouldn't have to get $30 million of VC funding, they could just get 15 million, and they wouldn't have to concentrate on telephony infrastructure at all, which they didn't want to do anyway.