Loraine Lawson spoke with David Gilmour, CEO of Tacit Software, on how illumio, a new type of Web 2.0 technology that helps connect the dots between unstructured knowledge, the people who have it and the people who need it-online and within the enterprise.
Lawson: The tech press has described your product as combining "a search engine with social networking." Is that fair? What can you tell us about how the technology works?
Gilmour: Actually, illumio combines attention management with what we call collaborative social networking. By attention management, we mean what people and information you should focus on right now. And by collaborative and social network, we mean collaborating with other people, but for the purpose of sharing information-not in the "MySpace" sense of making friends. We mean, "How does the group as a whole tap into itself as a resource, tap into the expertise that its membership might hold."
Nobody we know is trying to do this set of things. We're the only ones out there that make it possible if you're in a group of people to figure out who else in the group has that relationship or nugget of knowledge that can help you, without having to spam the whole group or have the group profile itself in some cumbersome way. We're trying to get the collective wisdom out of any group.
Lawson: How does the technology work?
Gilmour: It's pretty easy. You download illumio-and illumio is a download for a very simple reason: To work, it needs to become part of your world and have access to your computing environment and figure out what's going to be interesting to you without having to burden you. It allows you to join groups. In illumio, groups consist of two things: information feeds and people.
What illumio does on your computer is monitor all the groups you're in and all of the new content in any of your feeds, plus any questions or requests from other members of the group, and it monitors all of that 24/7 as long as you're on. Then, it figures out which of those articles and requests you should pay attention to and it tells you with popups, Let's say you join a group at work or a parents' association group at your school. If that group has a newsletter as its feeds, when those items pop up, illumio monitors it and sees which information will be of interest to you. And if the information is of interest, it pops up and tells you, "This is something you should pay attention to." This means you don't have to read the feeds for yourself.
The world's become full of feeds. The whole interesting question about Web 2.0 in the enterprise is, "Gosh if everybody is starting to put up blogs and feeds, how do I keep up? I can't read them all. I need something to be going to all these feeds on its own automatically and calling my attention to what matters to me. And - by the way - I'm busy. Don't ask me what's interesting to me. Figure it out for yourself."
That is the attention management problem to us-the big challenge we have now in a world where users are creating content, everybody's creating content, and everybody's drowning in content. There are others who are trying to help you filter content, but we're the only ones who do that automatically by figuring out for ourselves based on what you're running on your PC. We can look for patterns. We let you refine it, but we don't make you tell us-we figure it out.
To us, the place we're going that nobody has gone is that we're also focused on what people you should be paying attention to - not just information. So if somebody in your group-any sort of group-asks a question, is it one that you can answer? Should we even bother you about it? Should we recommend that you respond? To us, that's the next level problem: Just as you have to figure out which information to pay attention to, you also decide which people you want to engage with, and illumio figures out which person should respond to a question or a request and alerts them without them having to declare in advance what their expertise is.
That's a big deal because, in the past, products had you fill out a form and tell what interests you have. But there's no way to detect that you know me. Our relationship - that you sent me e-mail - is never going to come up. Yet, if somebody in the group says, "Does anybody know David Gilmour?" illumio sees that it's in your e-mail. It pops up and says, "Hey, Sally wants to know David Gilmour. We see you know him pretty well. Do you want to introduce him to Sally?" In that way, our job is to make connections between people for the purposes of collaborating and exchanging information and do that 100 percent automatically and privately.
Lawson: So, how are companies using this within the organization?
Gilmour: Between Tacit's ActiveNet and illumio, we have companies in a wide range of industries. Drug companies, investment banks, aerospace. Usually it's knowledge workers in those companies who need to find knowledge or expertise and they don't know whether it exists or not. We've had story after story of people who are developing products or they are talking to vendors and they go to our system and say, "is anybody else working on this?"
Lawson: It almost sounds like you're-I don't want to say capturing, because it doesn't seem like capturing in the traditional knowledge management sense. It sounds like this is a way for companies to tap into that Web 2.0 knowledge, which thus far is something companies haven't been able to manage, use or even find very well.
Gilmour: Capture is the wrong word. Knowledge management has historically been about capture. "Let's capture everything our employees know and put it in a server." We think that "capture" fails for a lot of reasons. There's no way to "capture" everything you know, plus people hold back information for their own political power.
Don't capture knowledge. Let it stay where it is and if someone needs something, connect them with the person who has that information. They'll figure it out: They'll know if they should swap PowerPoint presentations or have a conversation. Don't try to run around and "capture" all this content. That's the Web 1.0 model, where everything has to be captured in one point of access. Just let everything be where it is and if the collaboration is successful, then you can capture the resulting piece of content. Unlike the philosophy of capture, this approach can scale. You don't want the IT structure running around trying to figure out what everybody knows. That is an un-winnable war-that is a doomsday project.
Knowledge is power and most people see it as a negative, they see it as hoarding knowledge. I see it the other way. I believe if you give people respect, allowing them to control what they share, most of the time they will do the right thing. Let them control so they can use it wisely, so they can get credit. But if you just ask them to throw all their knowledge in a big pile, they will find someway to sabotage that. Whereas, if you allow them to control that information and reward them for sharing it, I think people are inclined to help, to share.
Lawson: How is illumio different from Tacit's flagship product, ActiveNet?
Gilmour: They're completely separate approaches to the same problem. ActiveNet is an enterprise technology that's deployed on enterprise computers behind a firewall and run by IT. It learns about employees by connecting to the e-mail servers. illumio is personal. illumio works independently on the personal computer of each group member. In illumio, it's different, because if you're in a group in illumio and you say, "Does anybody know how to put a silver plate on metal?" or "Does anybody know how to move a Saturn 5 rocket?" then that question goes up to the illumio service. illumio doesn't do anything with it-it's not like Google-the only purpose of our service is to faithfully deliver that question to the users when next they're online. Everybody calculates their own score of how well they matched the question based on all the private information on that computer, but nobody gets access to it.
Lawson: Aren't companies and individuals afraid that illumio will send out information about them?
Gilmour: We think the clever thing about illumio is it can work without doing that. So you didn't have to take our word for it, we became one of the first organizations to be certified under the Trustee program. You'll see us listed there. It will explain how they have been through our program and source code in detail and certify that we do indeed do what we say we do.
Users get it. There's no reason for illumio to give your information to somebody else. The whole purpose of it is to match you to stuff.
Lawson: But do enterprises get it?
Gilmour: They get it. We're very involved with the national security community. Privacy is not a liability; it's an asset. In the publishing system, you have to upload your information. In a knowledge management system, you have to upload your expertise. You have to part with information to give it to somebody else. But with illumio, you don't have to do that. You're downloading all the searches so you can know when one matches you. There's no reason to give information about you to anybody else.
illumio is so secure that it can work across security compartments, for example, in classified situations. Let's say you're working on Top Secret Project X and we're in a big group together. I can say, does anybody know about project X? Your computer gets the search, as everybody else's does. You match, because your computer knows about Project X. But what happens? You get a popup explaining that somebody else is seeking knowledge about this. But they have no way of knowing you received the popup-or that you even exist or that Project X exists - unless you choose to respond.
Contrast that with today's approach. In most matching systems, knowledge systems, the first step is to capture what you know and have you upload it into some system, which requires that you trust the person operating the system, because now you have to list your interests. In illumio's case, you never have to trust anybody, because illumio's operating privately, locally and independently on your computer. The searches come to you instead of you going to the searches.
With the enterprise version, you get a private group that stays on the system. All the service that's responsible for distributing the services knows is whom has the search been delivered to. When you do a search in an enterprise group-let's say it's Sunday night and you send a query out to your 500-person engineer organization. You see that 10 percent received the search, but on Monday morning, when everyone turns on their computer, the percentage goes up to 90 percent. You don't know if there's a match or they just ignored it. You just know the percentage that received the search.
Lawson: What's the cost for a single group?
Gilmour: If you want to just use it publicly with no wall, it's free, but if you want to cut it off and have your own space to use it, that's what you pay for. You don't pay for a public group. You pay for a private group. It's $5 per person per group. For enterprises, we discount that for multiple groups so it's not that expensive. The fun stuff is when you join a group. The group comes with feeds. So if you join a group with stamp collecting, you're going to put in the feeds, and if somebody joins your group, they get all the feeds.