Raising The Ante
omputer gamer Gilman Louie now plays for keeps.
Former chief of online projects at toy maker Hasbro, inventor of the popular Falcon F-16 flight simulator and importer of the hypnotic Soviet computer puzzle Tetris, Louie now ferrets out emerging technologies and invests in them for the CIA. He is the most visible of a new breed of government buyers. They're using venture capital to lure technology innovators into the federal market they once shunned. Louie runs In-Q-Tel, a 3-year-old nonprofit "venture catalyst" created and funded by the CIA to place bets on technologies likely to succeed at the agency-as well as in the commercial market.
Louie's at the leading edge of a desperate push to speed the latest, greatest gizmos, gadgets and software into the hands of federal users. Especially since Sept. 11, legislators and the Bush administration are intent on driving high-impact technology into government-especially the Defense Department and intelligence agencies. If that means powering over, past or through traditional contractors, procurement methods and people, so be it.
The Defense Department has taken to pleading publicly for anti-terrorist technology. The first such call, in October, netted more than 12,000 proposals. A second was issued in March. This year's $159 million Defense Advanced Concept Technology Demonstrations program is larded with homeland security and counterterrorism measures. The 7-year-old program seeks to bypass the acquisition process, quickly prototyping developing technology and getting it into use. In December, Defense launched a dozen rapid improvement projects to test ways to implant new technology within 18 months.
Outside Defense, the White House Office of Homeland Security dedicated four staffers to meet with industry and weigh technology proposals. Sen. John Edwards, D-N.C., announced in February his plan for a national test bed to evaluate new technologies for the war on terrorism. In April, Rep. Tom Davis, R-Va., announced he wants to create a program at the Office of Federal Procurement Policy to help take advantage of homeland defense and anti-terror technology innovations.
Well before Sept. 11, U.S. intelligence agencies realized they had fallen behind the technology curve. It was a difficult, embarrassing admission for the folks responsible for intelligence satellites, the U-2 spy plane, the SR-171 Blackbird reconnaissance aircraft and a host of other groundbreaking gear. But times had changed. The Soviet Union's collapse in the early 1990s removed the overwhelming rationale for spending any amount to gain a technological edge. That time also saw the rise of the World Wide Web, the beginning of the information onslaught, and the start of an era in which industry vastly outspends government on research and development, especially in information technology.
"The Russians were a simple threat. It was easy to come up with focused, targeted solutions," says Michael Crow, executive vice provost at Columbia University and In-Q-Tel board member. "When you are looking at an enemy that can vaporize your civilization in an instant and somebody says, 'I need a handmade satellite and I'll pay anything,' well, the world doesn't work that way any more. Now, [enemies] have an advantage: They're using the same technologies the market is using and the Agency is not accessing the market. In certain areas of technology, the private sector is bigger than the government and can move faster." Crow was named Arizona State University president on March 29.
Intelligence visionaries saw that the speed of technological development, industry's spending lead, and the loss to downsizing of considerable tech talent, meant government had little choice but to try to harness the innovative energy of private firms. One of those seers, Ruth David, had come from Energy Department research partner Sandia National Laboratories to run the CIA's science and technology directorate in 1995. She found the situation dire. "I had to ask for Internet access on my desktop," David says. "Not tying the Internet to the internal network, I agree with, but not having it available on an independent network? People in the intelligence business need to live in the world of information."
David, now chief executive officer at Anser, a nonprofit research institute in Arlington, Va., found the Agency's technology team isolated and far too reliant on a handful of large, CIA-savvy contractors. "If their job is to infuse the best technology to enable the mission, then they need to have access to it, and that comes in communications with the people developing the technology," she says. "The traditional contractors are not the hotbed of innovative IT."
Elsewhere in the intelligence world, the National Reconnaissance Office, which runs the U.S. spy satellite network, already had begun moving beyond traditional buying methods and suppliers. "There was a small [NRO] group that realized in the late 1980s that government was not driving technology, but had become just another customer," says Mark Lister, senior vice president for government operations of Sarnoff Corp., a Princeton, N.J., research and development firm, and managing director of Rosettex Technology and Ventures Group. "To capitalize on industry investment, they realized, government would have to become a player. NRO had to. . . form relationships with companies in the commercial world."
In 1987, struggling with the magnetic tape it used to store a huge volume of images and other data, NRO reached beyond its contractor comfort zone, bringing in 3M, then a leading recording technology firm with little government experience. 3M solved the problem and NRO got the firm to lead a consortium of corporate and academic partners in creating the National Media Laboratory to take further advantage of commercial advances in information storage and retrieval. Three years later, NRO created a National Information Display Laboratory, hosted by Sarnoff, to help solve problems related to sharing and displaying images and data. In 1992, the labs were folded into the National Technology Alliance, which is dedicated to finding and developing commercial IT products for use by Defense and intelligence agencies.
In February, the National Imagery and Mapping Agency, executive agent for the alliance, abandoned the labs and awarded Rosettex Technology and Ventures Group a contract worth up to $200 million over five years to be the National Technology Alliance's In-Q-Tel. The company is charged with leading a team of experts sniffing out best-of-class technology in imagery, cartography, digital processing and analysis, and digital infrastructure. Rosettex is a joint venture of Sarnoff and SRI International, a nonprofit research institute in Menlo Park, Calif. Both firms license technologies, and they've created 26 new companies between them, according to Lister. Rosettex will invest profits from the technology alliance contract in a venture fund, which, like In-Q-Tel, is intended to support technologies useful to government and the commercial market.
SURFING SILICON VALLEY
The government's procurement culture is just too suspicious of business needs and practices to embrace flexibility. Business, especially the fast, high-rolling technology sector, distrusts government's contracting methods and finds its oversight onerous. What's more, when Ruth David went technology prospecting in California, high-tech startups bluntly told her the CIA's market share just wasn't big enough to bother with. "The government still has not come to terms with not being a big player in a market it helped to create," she says. As David sought ways to marry the Agency and the Valley, control of the CIA passed in 1997 to George Tenet, who brought along investment banker A.B. "Buzzy" Krongard, first as his counselor and later as executive director. A tough-talking former Marine infantry officer with the polish of Princeton, Krongard had been CEO of Alex Brown Inc., the seventh-largest bank in the country and an early investor in Microsoft and AOL. Krongard and Tenet embraced David's effort.
"I immediately saw the attractiveness of the idea," Krongard says. "Alex Brown had the reputation of being the No. 1 technology bankers in the country. Consequently, we had met, interfaced, romanced, cajoled, begged for business-whatever-a lot of different people. Their attitude was such that they were not gonna put up with the typical government approach to doing business, and yet [that was] where the cutting-edge stuff was being performed."
Tenet decreed the CIA had to "swim in the Valley," and a new relationship between the Agency and tech firms was born.
Jeffrey Smith, In-Q-Tel's legal counsel at Washington law firm Arnold & Porter and formerly CIA general counsel, worked with Krongard to build a board of outsiders to fill in the outlines of David's vision. "The CIA didn't quite know what it wanted-it wasn't a federally funded research and development center, it wasn't procurement, it wasn't in-house," says Louie. "So they went to Norm Augustine, not as [former] Lockheed CEO, but as a maverick, and said, 'Help us put together an all-star, killer board of directors to wrestle with the problem and create a model.'"
That model mimicked venture capital firms, right down to their equity investments in startups and new technologies. "From the beginning, we knew equity drove the IT market," Smith says. "Equity was the genius, the energy. It was the only thing that would give us credibility in the open rating system of Silicon Valley."
But investments were to be just one tool at the organization's disposal, a tool to be used sparingly-most venture firms are funded at $100 million or more, more than three times In-Q-Tel's annual budget. So In-Q-Tel invests cautiously; the board must approve any investment of more than $250,000. To allay legislators' and the CIA's fears that they might lose control of In-Q-Tel if investments really paid off, all proceeds must be reinvested in further technology research and acquisition. "Our goal is not to collect a bunch of stock," says Louie. "We're using the approach to get products and services that are going to make a difference to the world of intelligence. What happens to the equity is a byproduct of the style with which we engage the company."
Investments buy the CIA access to technologies still in the womb. In-Q-Tel's board saw that "if the CIA could be involved at the beginning as a technology was being developed for commercial applications, the company could be modifying it so it would be useful to both industry and the CIA," Smith says. "As soon as it hit the market, the CIA could buy and use it and be as current as anybody in the world."
As the board developed a model, agency insiders canvassed the CIA's technology needs. The board then shopped the list, or "problem set," among research organizations, academics and tech companies, asking, "'You got anything? This interest you? We've got $30 million and the CIA as a potential customer,'" Smith says. Annually updated problem sets now determine In-Q-Tel's direction.
As for choosing a CEO for the new organization, the board "wanted a 30-year-old with a ponytail," Smith recalls. Board member John Seely Brown, who had run Xerox's famed Palo Alto Research Center, recommended hiring a computer gamer. Some of the very best talent in the Valley was in gaming, he said, because gamers have to push technology to its limits.
At the time, Louie, who had created his first computer game company while in college and had just sold his latest company, MicroProse, to Hasbro, was having a blast in the toymaking business. He recalls thinking, "It's the movie 'Big' for me. I'm Tom Hanks." Invited by a business magazine to participate in a mock dogfight put on by an Atlanta flight school, Louie was "shot down" by corporate headhunter Randy Jayne, who was working for In-Q-Tel's board. Jayne gave Louie a concept paper about the CIA's plans and pressed him to interview for the CEO slot. But Louie was reluctant to leave Toyland. "My father was a Navy planner and estimator for construction projects, a loyal government employee," says Louie. "I always was interested in government, but it didn't pay, and I believed you couldn't get anything done in government."
What's more, Louie, a Democrat who had demonstrated against California ballot initiatives favoring English-only classes and opposing affirmative action, was not a likely spook. But the CIA paper intrigued him and convinced him the Agency wanted to change. "I'm an American-born Chinese, fourth generation, a 'banana,' " Louie says. "In the circles I hung out in, there was a lot of discussion about how the U.S. misread the climate in China and couldn't understand how Chinese youth had become more nationalistic. They were the Internet generation and the U.S. thought they would be sympathetic and not want to fight over Taiwan. But that was not the case. So I thought that if the intelligence services could do a better job of understanding the other side it would lead to less mistakes and better policy.
"I decided to take the job because I was convinced that the Agency was serious. With these people, it's a game for keeps," Louie says. The board was delighted, despite Louie's leanings. "Gilman knocked the socks off the interviewers with grasping what we wanted," says Krongard. "So much of the technology is in California, and Gilman was very well connected out there."
Q'D INThe New York TimesThe Washington Post
"The coverage worked in our favor," Louie says. "The high-tech companies in the Valley are run by people who were kids in the 1970s and 1980s. They thought Bond and his gizmos were cool and if they had cool Bondish technology, they'd have a big customer in the CIA." The homage to gadget master, "Q," in In-Q-Tel's name, multiplied the Bond effect, surpassing the wildest hopes of In-Q-Tel supporters. "We assumed the Valley would be intrigued by the CIA, but reluctant to deal with it," says Smith. "To our delight, that was not the case. The smarter ones recognized an opportunity for them. When they worked with the CIA technical people they discovered they were enormously talented. Plus, they recognized the imprimatur of the CIA is a useful selling device: If it's good enough for the CIA, it ought to be good enough for you."
The reception inside the CIA was markedly cooler. Fiercely protective of their own projects, CIA program chiefs resented having their budgets reduced by the $28 million it cost to start In-Q-Tel. They also doubted that a group of slick, geeky outsiders could possibly find technology on the Agency's behalf. A congressionally mandated status report found that In-Q-Tel needed to enlarge its PR campaign to include the CIA. The vehicle for that marketing effort and for getting In-Q-Tel's finds adopted within the Agency is the In-Q-Tel Interface Center, known as the QIC.
The center, made up of 13 to 15 widely experienced, well-respected CIA staffers, is responsible for overall planning and management of In-Q-Tel's relationship with the agency. "We had to come up with a mechanism to lock the two together to avoid too much of a culture shock between In-Q-Tel and the Agency, [so] we created QIC," says Krongard. "QIC's a broker between In-Q-Tel and the Agency. They do every thing from say, 'That's not true,' if they hear a rumor about In-Q-Tel, to sifting out the ideas. Somebody has to take a corporate view-If we're going to spend dollars and time, where is there a big payoff for the entire organization?"
In-Q-Tel's 40 employees, along with the interface center staff, have become adept at identifying solutions and getting them into the Agency. They have engagements of various types with 20 firms and investments in 14. About 80 percent of In-Q-Tel's $35 million annual contract with the CIA goes into project funding, investments and the costs of transferring technologies into CIA programs. Unlike a typical venture capital fund, In-Q-Tel measures its success not by return on investment, but by return to the CIA in the form of technologies adopted. Since 1999, In-Q-Tel has delivered 19 technologies and demonstrations of 6 others, mostly in the fields of knowledge management, enterprise intelligence, and data classification-all methods of distilling knowledge from masses of different types of data from multiple sources.
In-Q-Tel employees tap into the high-tech deal flow by working the room wherever techies and venture capitalists hang out. Increasingly, they collect leads from other venture capitalists seeking to vet promising firms by piggybacking on In-Q-Tel's fiendishly rigorous technical due diligence. In-Q-Tel has become a sought-after investment partner as well, which nets early intelligence from other venture capitalists. Since Sept. 11, patriotism has swelled the proposal stream-more than 60 percent of the nearly 2,000 proposals In-Q-Tel has received since 1999 came after the attacks on the Twin Towers and Pentagon.
MORE THAN MONEY
Take In-Q-Tel's arrangement with Mountain View, Calif.-based Stratify. The Stratify Discovery System classifies and manages unstructured data-documents, e-mail and the like. When Stratify began its mating dance with In-Q-Tel, the firm had raised more than $35 million since its creation in 1999. "Whatever few million [In-Q-Tel] brought wouldn't make any difference," says Nimish Mehta, Stratify president and chief information officer. "The attractive part of the relationship was their knowledge of the government market, their position, the introductions," Mehta says.
Zaplet, a Redwood Shores, Calif., maker of collaborative tools that allow software applications to travel by e-mail and be automatically updated by each recipient, wasn't looking for money either, says co-founder David Roberts. "We had just finished a $90 million financing round." So instead of investing directly in Zaplet, In-Q-Tel bought the firm's product and services and credit to buy stock down the road at a set price. "We wanted to say, 'We are interested in your product and services and we will help you build a government market, but we want to understand what's in the heads of your investors-we want a strategic-level engagement,'" Louie says.
The Stratify and Zaplet engagements are among the two-thirds of In-Q-Tel deals that Louie deems "evolutionary"-new ways of applying known technologies to solve new kinds of problems. Another 10 percent are revolutionary. They are new science with a hot new market and no competition. An example is In-Q-Tel's investment in La Jolla, Calif.'s Graviton, which makes Web-based wireless sensor monitoring and control systems that could, for example, allow an electric company to measure the temperature of electrical transformers and alert the customer if one overheated. Such sensors could have myriad intelligence uses for monitoring equipment, communications and the like. The rest of In-Q-Tel's deals involve mature technologies, such as System Research and Development's Non-Obvious Relationship Awareness soft- ware. Popular with the casinos in Las Vegas, where SRD is based, as well as among retailers, NORA surveys a company's collections of data to identify relationships that could point to fraud and collusion among employees, clients, suppliers and outsiders. The software has obvious applications for detecting terrorist sleeper agents and scrubbing airline reservations. SRD CEO Jeff Jonas says he is using In-Q-Tel's money to add a secret "ultimate feature" to NORA. Especially attractive to the CIA, the feature should make NORA even more attractive to corporations. "To my shock, everything we're doing, every feature valuable in the federal space, is just a bigger problem that also goes on in corporate America," Jonas says.
But not everyone is keen on In-Q-Tel proliferation. "It's not in In-Q-Tel's interest, because right now we've got the field to ourselves," says Krongard. "So we are cooperative and we tell everybody the good, the bad and the ugly, but if they think I'm trying to sell them on having [an In-Q-Tel] they're wrong." Lister worries that agencies won't fully appreciate the risks involved in betting on untested firms and products and that creating more funds will dilute potential technology gains. "If NRO and [the National Security Agency] and NIMA and the Army and Air Force and Navy and Marines all do In-Q-Tels, the competition is going to be enormous. And I question how much return on investment you get as a community," he says.
Last year, the Army Science Board said the Army, and federal agencies in general, "should avoid owning a minority interest in private companies." The panel "disfavors the In-Q-Tel model of making anticipated return on investment an evaluation factor" because "it detracts from the foremost objective of selecting vendors with the potential to develop technologies necessary for the objective force." The Army currently collects little more than $400,000 a year in royalties on products developed with industry, so the board also doubted the existence of commercial markets for technologies in which an Army In-Q-Tel might invest.
"I went to the Army and said, 'Why not spend your money through Rosettex?'" Lister says. "Do your R&D and we'll put our fee in the venture fund and Army leaders will be gaining some knowledge." Lister already is working with In-Q-Tel and has invited Louie to sit on the board of Rosettex's venture fund. "Gilman and I are interested in working collaboratively so people who come to him with ideas that are too early-stage for him and need prototyping he will shoot to me. He'll also see stuff he could pick up by sitting on the Rosettex board," Lister says. "Our models are complementary: I can't compete with him because I don't have enough money. He will not compete with me because the stuff Rosettex invests in is too early in its development for him."
If In-Q-Tel and its look-alikes continue to cooperate with one another and other venture capital funds, they may succeed in speeding the flow of fast-evolving, useful technology into federal agencies. That won't quiet critics who believe government has no place in business, but it might help intelligence agencies better grapple with the multiple threats and unfathomable welter of information they confront in the post-Cold War era of terror. "We live with fear in this business," Krongard said in an interview 10 days before Sept. 11. "One of our biggest fears is that something happens today, and when we do the autopsy we find that two weeks ago we had it, [but] we didn't know because it was buried in something else that wasn't getting processed or we lacked for a Pashtun speaker or who knows what."
A 'KILLER BOARD'
- Lee Ault, chairman of the In-Q-Tel board; former chairman and CEO of Telecredit Inc.
- Norman Augustine, former chairman and CEO of Lockheed Martin Corp.
- John Seely Brown, former director, Xerox Palo Alto Research Center
- Howard Cox, general partner, Greylock; chairman, National Venture Capital Association
- Michael Crow, president, Arizona State university; former executive vice provost, Columbia University
- Stephen Friedman, senior principal of Marsh & McLennan Capital Inc., and former chairman of Goldman Sachs and Co.;
- Paul Kaminski, president and CEO of Technovation Inc., senior partner in Global Tech- nology Partners, and former undersecretary of defense for acquisition and technology
- Jeong Kim, former group president, Lucent Optical Networking Group, part of the Lucent Technologies Group, and founder of Yurie Systems
- John McMahon, former president and CEO of Lockheed Missiles and Space Co., and former deputy director of central intelligence
- Alex Mandl, former chairman and CEO of Teligent and former president and CEO of AT&T
- William Perry, former secretary of Defense
- Internet Search and Discovery
- Web crawling, indexing, ranking
Information Security and Privacy
- Adaptive threat detection
- Network privacy/anonymity
- Digital rights management
Enterprise Knowledge Management and Visualization
- Enterprise search/retrieval, indexing, access management, personalization
- Data warehousing/mining
- Collaboration environments
Geospatial Information Services
Distributed Sensing/Data Acquisition