1 July 2002
In-Q-Tel, Investing In
CIA Unit Scours Country For Useful Technologies
By Shannon Henry
Washington Post Staff Writer
Monday, July 1, 2002; Page E01
Like "Q," the gadget-maker who keeps James Bond perpetually ensconced in the latest high-tech gear, Gilman Louie is looking for technologies and ideas to give American spies an edge.
Louie is the founding chief executive of In-Q-Tel, the venture capital unit of the CIA that -- no kidding -- named itself after the movie character. The group, created in 1999, has made about a dozen investments in technologies that could potentially be used in information gathering and analysis of America's enemies.
It was always a controversial concept -- the U.S. intelligence community openly investing in pieces of commercial technologies. Why wouldn't it just buy a technology outright or develop it themselves? But the benefits of having In-Q-Tel's fingers in a lot of commercial technology pies were demonstrated after Sept. 11, when In-Q-Tel found itself a go-to group. Suddenly, it was necessary to scour all parts of the country for technologies that could be used for counterterrorism and homeland defense. A vital concept could come as easily from a Las Vegas entrepreneur intent on catching crooks in casinos as a government researcher toiling away in a laboratory.
Government agencies came to the organization for technological advice and expertise. Companies' executives began deluging In-Q-Tel's Rosslyn offices with business plans -- two to three times the number before the attacks, about 150 a month.
In no time, In-Q-Tel became a sort of anti-terrorism matchmaker, introducing those with problems to those with high-tech detective abilities. Several federal agencies, including the Army, the Navy, the Defense Department and the National Aeronautics and Space Administration, also are studying whether it is possible to replicate the In-Q-Tel model or partner with the enterprise in some way, Louie said.
"Everybody is watching and trying to figure out how it will fit into their culture," Louie said.
In-Q-Tel's development marks a stark departure from the way government research and development often is conducted, often in secret and distinct from the commercial sector.
Louie spends half his time in Washington, meeting with politicians, lobbyists, government administrators, local technologists and financiers. He spends the other half in Silicon Valley, where he networks with venture capitalists and tests out ideas for new technologies.
"The fun part is going into [CIA headquarters at] Langley and talking to national security people and giving them tools to get their jobs done," Louie said.
So what is he looking for these days?
Because the organization is part of the CIA, after all, Louie will not get into specific details of what he is looking for or why. But In-Q-Tel is focusing much of its energies on tracking terrorists, finding links between criminals and even guessing what they might do next.
Over all of these efforts is the apparent lack of information sharing and some crucial lapses in analysis that took place before Sept. 11, when different parts or offices of the CIA and the FBI were not able to pursue leads that might have led to the hijackers. Technology, or the lack of it, in FBI field offices has been cited frequently as an issue in these failures.
Solving such "knowledge management" problems is top on Louie's list, particularly using technology to see patterns and massive amounts of data. "How do we make information overload our friend?" he said. "Can you be predictive?"
He also is looking closely at distributed computing systems -- how networks can be linked and how people in different places can share information.
Louie said he gets many of his tips for promising technologies from traditional venture capitalists. "They have my laundry list, and I have theirs," he said. He found one investment, Attensity, a natural-language-processing company in Salt Lake City, when a venture capitalist suggested he take a look.
He compares Attensity's software, which extracts common threads of information out of documents, to a kind of high-tech sentence diagramer. In-Q-Tel often will invest in companies like Attensity, which have created products for the commercial market that might also have government application.
Although it would not say exactly what they were doing with it, In-Q-Tel is now piloting Attensity's software, which was created to comb through vast databases to help large corporations understand more quickly what their customers are telling them. For example, do customer complaints about a certain toaster suggest design changes, or is a recall warranted? Todd Wakefield, chief executive of Attensity, compares those customers with field agents of a government office.
"The intelligence community's issue isn't gathering data, it's analyzing it," Wakefield said. "Their biggest problem is free-form text."
For Wakefield, In-Q-Tel's attention was a company-saver. A huge commercial deal that was to be signed Sept. 12 fell through after the attacks. He thought about getting on the General Services Administration schedule to sell to the government but figured it would take too long to get established. In-Q-Tel's interest helps build credibility.
David Gilmour, chief executive of Tacit Knowledge Systems in Palo Alto, Calif., said it had never occurred to him to market his software to the government. His customers were huge pharmaceutical and aerospace firms that wanted their employees, often working in widely dispersed satellite offices, to share information better by understanding who was working on what. In-Q-Tel heard about Gilmour and invested $1 million, and is actively using Tacit technology.
"It translates directly to the problems we read about every day in the headlines" about the FBI's management problems, Gilmour said. He said so much important information, such as informal notes, is not easily searchable online.
And then there was the Las Vegas inventor who created "link analysis" for casino owners to use to spot illegal activity among gamblers and dealers. Louie, who talks quickly and excitedly, said: "Casinos want to catch these guys in real time." And so do spies.
So Systems Research and Development of Las Vegas became one of In-Q-Tel's investments. SRD claims to be able to show up to 30 degrees of separation from any person -- a banker in Tucson knows a broker in Dallas who knows a hotel owner in Miami and so on. "We deliver an investigative clue," said John Slitz, chief executive of SRD. Company founder and chief scientist Jeff Jonas says the software in three minutes can find if a person is a known terrorist or linked to one by entering a name in a database.
About 20 percent of SRD's current business is with the government, and company officials hope to see it jump to 50 percent in the next year.
In-Q-Tel is building an extensive portfolio, but there are still many challenges, not the least of which is the shock of the new in the culture of the CIA. Louie said it was difficult to convince everyone, commercial and government types alike, that investment in computer infrastructure, not just spying gadgets, is necessary.
"It's a real funding issue," Louie said. Right now, In-Q-Tel has a $30 million annual budget and 45 employees. But the venture capital model allows In-Q-Tel to stretch those dollars across many technology bets.
In-Q-Tel judges investments differently from most venture capitalists, who look toward a company "exit strategy," usually an initial public offering or a merger. Louie said obviously they want to invest in successful companies that will not disappear, but candidates more importantly need to show a technology that works, one that can solve a problem at a government agency, or even address a problem no one had perceived before. Through these investments, In-Q-Tel becomes a special customer with a greater knowledge of how the technology can work for government agencies, and encourages innovation that might otherwise fade away. Louie said In-Q-Tel does not ask for exclusivity arrangements because the organization's whole point is to leverage commercially available technology.
"We are casting a big, high net and we are talking to everybody and anybody," Louie said. "We can't afford to get it wrong."
© 2002 The Washington Post Company