Most of the truly valuable information that an organization possesses is ac-tually floating around in its electronic-mail system. The problem is that there is no viable way to easily access that data.
But now along comes Tacit Systems, which has developed a set of tools that allows an organization to mine e-mail for relevant data. In an interview with InfoWorld Editor in Chief Michael Vizard, Tacit CEO David Gilmour talks about why he thinks his new approach to knowledge management is going to finally solve an age-old problem.
InfoWorld: Why has e-mail been such an elusive asset when it comes to knowledge management?
Gilmour: I think there have been two main barriers to trying to see e-mail as something different and something more valuable. The first has been technology.
What's required is a different kind of profiling than anyone has ever done before. That profiling ability has to involve a sense of time and not just how a group of documents or messages relate to each other. It also needs to know what order they came in and how the attention of a given user to a particular concept has gone up and down over a period of time. The other issue is privacy.
E-mail is a subject that arouses a lot of strong feelings. Even though we all know the legal ownership of e-mail in the United States lies with the corporation ... users often feel that if an enterprise were to simply monitor e-mail unilaterally, it would be a tremendous invasion of them personally. So we needed to take a completely new approach to the privacy problem. In particular, we needed to show that it was actually in the enterprise's interest to give the user a safe haven or a private area on the server where information could be captured and assembled. This then allows [users] in a very efficient way to decide how others should perceive them and how they should relate to others.
InfoWorld: How is that actually accomplished?
Gilmour: The idea is that [all users have] a profile, which is the way they appear to other people and the way they can be searched and browsed. We make it possible for everybody in an enterprise to have a profile because we automate the process by which that profile is created and maintained. One of the problems with previous attempts at knowledge management has been the amount of user involvement. Users had to remember to contribute information to a repository and [to] maintain a profile. In the real world, the very users you most want to participate have the least amount of time to do that. What we do is give them a means to participate that doesn't require a lot of time or effort yet gives them a sense of control. Users can manually enter data, or by right clicking on any file they can update the profile appropriately. In the fullest scenario of automation, users can allow us to get copies of their outbound e-mail. We then carefully factor out that part of the e-mail that the users themselves sent from their fingertips and use that to update their profileInfoWorld: How do individuals benefit from doing this?
Gilmour: The first benefit is that because the system coaches you on how to have a public profile that others can search and browse, suddenly you become known to other people, and you become reachable. You become understood for who you are, what you do, what you know, which means others can tap your knowledge when they need you to get their jobs done. You can get ahead in the enterprise; you can advance your career and show that you're a success in your enterprise.
For you, personally, it means for the first time you can project to others a very real moving picture of who you are. To the extent that you choose not to reveal things about yourself, they stay in the private profile, because by default, everything that we learn about you starts in the private profile. It is only when you choose to have [information] become public that it's revealed to others. If somebody else is looking for somebody that has your knowledge or skills, the system will actually broker a connection between you and that person if you want it to. So it's also a way of putting an ear to the ground everywhere all the time in your enterprise and knowing when you should inject yourself into a situation. But you're never in a position where something is revealed about you without your wishing it.
InfoWorld: How complicated is it to build this?
Gilmour: We've worked very hard to be non-invasive to the e-mail infrastructure and leverage it wherever it exists. We do that by requiring on the server side that a mailbox be created for our product. Then when end-users install the knowledge mail-client software [and] they have Microsoft Outlook or Lotus Notes, we automatically detect that and get copies of their outbound messages, assuming they've enabled that. We're not modifying the mail system, we're not modifying the servers, and we're not increasing the traffic.
InfoWorld: Based on what you just described, does this product really fit the definition of knowledge management?
Gilmour: We don't think of ourselves, first and foremost, as a knowledge-management application. In fact, we say we deliver knowledge without the management. And the reason is that the knowledge is what people want; the management is mostly cost and trouble. Our focus is on automation on the theory that to be valuable to an enterprise, everybody has to participate. For everybody to participate, it has to be automated. For it to be automated, it has to be based on something everyone has, which is e-mail. And to be based on e-mail, it has to have a very radical, new solution to handle the privacy issue. What we're about is transforming e-mail, not simply focusing on the people who think they're doing knowledge management.
InfoWorld: Are there other applications beside e-mail that you need to tap into?
Gilmour: We allow IT departments to integrate our profiling engine with other enterprise applications.
InfoWorld: Are you also compatible with network directories?
Gilmour: The way the relationship works with directories is that when we initially form a profile on somebody -- perhaps from a published source -- a document is created in the repository. We then can create a new account for that user, and we automatically reconcile that account with the LDAP [Lightweight Directory Access Protocol] information in the directory, so that the user's name, ... location, [and] any of the other LDAP parameters that are served up by the directory server are passed through to us. That way the user doesn't have to maintain information in two places.
InfoWorld: Technically, it sounds like you may have the problem solved. But culturally, these types of products have historically been a tough sell. How will you overcome this?
Gilmour: We believe that in every enterprise there's a range of users from those who are very ready to share knowledge, to those who are very skeptical.
Our goal has been to address everybody along the spectrum. If you install our product and do absolutely nothing, don't publish any information about yourself when the system suggests it, don't go to your Web page and so forth, we will add value to your life by simply sending you e-mails when situations arise that you should be aware of.
InfoWorld: So at the end of the day, how would you define the relationship between Tacit and e-mail?
Gilmour: Our primary role isn't to change e-mail, it's to turn e-mail into an asset that makes people work better and increases [e-mail] value. We do also do that, for example, by allowing people to take a draft e-mail message that hasn't been sent, have the system analyze it on the fly, and produce back a list of all the people who should receive that e-mail message. That's important because in many organizations, e-mail is over-distributed to people who don't need or want to receive it. ... It's also under-distributed to people who should receive it, but because they didn't receive it, they miss out on some critical piece of knowledge.
Tacit Knowledge Systems Inc., in Palo Alto, California, can be reached at http://www.tacit.com.