Mercury Rising


For most of the history of computing, most of the industry's collective weight has been focused on automating machine-to-machine processes. Now, to be sure, we have created wonderful personal productivity tools, but any process that requires mixing machine and human interaction on a regular basis has been difficult to create.

That's why most of the enterprise application integration (EAI) scenarios that have been deployed have primarily focused on machine-to-machine applications. In recent years, EAI vendors have begun to embark on creating EAI platforms that could better integrate people into a complex business process. But as each of the EAI vendors has approached this challenge, they have created a proprietary approach that can't easily be replicated across multiple business-process integration platforms.

That's what makes Project Mercury at Avaya Labs interesting. The folks at Avaya Labs, a descendant of Bell Labs, have always been focused on communications. But with Project Mercury, Avaya is venturing into business process integration. The project, led by research scientist Ajita John, is developing a Web services-based architecture that makes it easier to plug a human being into an automated set of business processes by essentially creating a series of parameters that will kick off alerts about changes and exceptions to business rules surrounding any set of business processes. In essence, any change that exceeds a certain threshold can trigger a request for human intervention.

Most recently, Avaya Labs has extended Project Mercury by linking it to a separate GhostReporter project led by research scientist Xueshau Shau, who has come up with a way to forward alerts generated by Project Mercury to any mobile device.

The reason people should care about these two projects comes down to the basic argument over the value of IT. We all know this issue is on a lot of people's minds these days, given the tepid growth in IT. But the real reason that the average business person subscribes to this point of view is because IT systems provide information after the fact. That means that for all the promise of IT, it's still viewed as simply a more efficient way to find out about something that has already occurred. What business people want is an IT capability that tells them about events in a way that allows them to alter those events before they become a fact of life.

Now that doesn't mean one alert is necessarily going to change the world. But that one alert can most certainly identify the beginning of a bigger trend that could either be minimized or accelerated.

As an industry, we're a ways off from routinely providing this kind of capability inside our business applications. But clearly, that's what the average business executive really wants and needs if he or she is ever going to truly see IT as a strategic, rather than tactical, business asset.