Incorporated: 1980 as IntelliGenetics, Inc.
Sales: $25.6 million (2001)
Stock Exchanges: NASDAQ
Ticker Symbol: INAI
NAIC: 511210 Software Publishers
Once a pioneer in artificial intelligence (AI), Mountain View, California-based IntelliCorp, Inc. now offers consulting services and software to Global 2000 companies. With offices in the United States and Europe, the company develops and supports a software suite of eBusiness and Customer Relationship Management solutions. With these tools, its consultants are able to help clients integrate computer systems that have differing data structures and other technology features, thus allowing the programs to share information and interact in a cost-effective manner. For instance, a client may have a number of different customer channels, including the Internet, wireless, contact centers, and field sales representatives. IntelliCorp helps to successfully connect these sources to back office processes, such as order management, order fulfillment, manufacturing, human resources, and accounting. Intellicorp products are designed to be used in a wide range of industries: apparel, appliances, chemical, construction, energy, financial, high-tech, insurance, manufacturing, media, pharmaceutical, and telecomunications. IntelliCorp’s major customers include Boeing, General Motors, and Hewlett-Packard.
Birth of Artificial Intelligence: 1950s
The leading figure behind the creation of Intellicorp was scientist, author, and visionary Edward Albert Feigenbaum. Born in 1936 in Weehawken, New Jersey, the son of an accountant, Feigenbaum grew up fascinated by his father’s hand-cranked adding machine. During the early 1950s, he attended the Carnegie Institute of Technology to study electrical engineering and became familiar with the new field of artificial intelligence (AI). The concept was espoused in a paper, “Can a Machine Think?,” written by British mathematician Alan Turing in 1950. Two Carnegie professors, Herb Simon and Allen Newell, then advanced the idea and created a Logic Theorist program that centered on the problem-solving process. According to a Forbes profile, “Feigenbaum recalls Simon announcing to the class one day that he and his colleague Allen Newell had invented a thinking machine. To help explain, Simon handed out a manual for an early IBM computer. Feigenbaum, then 19, stayed up all night reading the manual. “By the next morning, when the sun came up, I was born again,” he says. “Here was this beautiful modern electronic version of what I was so intrigued with as a kid, that little mechanical calculator of my dad’s. And furthermore, Herb Simon had said, ’I’m going to show you how to make them think.’”
Because there were no computer science programs, Feigenbaum earned his Ph.D. in industrial administration at Carnegie, becoming involved in research that enabled a computer to model the rote learning of nonsense syllables. After teaching at the University of California at Berkley, he joined the Stanford faculty in 1965 as an associate professor of computer science and director of the school’s computer center. Feigenbaum now began to make his own contributions to AI. While working with a group developing an esoteric system that could analyze unidentified chemical compounds, he realized the breadth of knowledge a chemist drew on to form decisions, a lot of which resulted from educated guesswork. Any computer system would likewise have to include a large database of rules in order to make similar decisions. Feigenbaum’s group began to call these types of comprehensive programs “expert systems.” The person who observed and codified all aspects of a particular discipline for a computer program was the “knowledge engineer.”
Incorporation of IntelliGenetics: 1980
Feigenbaum worked with other departments to develop expert systems in the sciences, such as organic chemistry and molecular biology, as well as medicine. He became a leading champion of expert systems, and AI in general, and by the late 1970s began to team with his graduate students and colleagues to transfer the new discipline to the business world. While computers had already automated clerical tasks such as accounting, AI-enhanced machines were expected to one day assist executives in making decisions. In September 1980, along with Douglas Brutlag, Laurence Kedes, and Peter Friedland, Feigenbaum incorporated IntelliGenetics, Inc. In keeping with much of the work Feigenbaum did at Stanford, the company’s first product was a package of computer programs for research in recombinant DNA. By 1982, IntelliGenetics generated little more than $100,000 in revenues with its narrow applications, and posted a net loss of more than $500,000. In 1983, the company introduced its Kee System, a general purpose AI tool that had the potential to develop knowledge-based systems with commercial and industrial applications. The company went public in 1983 at $6 a share, then a year later changed its name to IntelliCorp, while retaining the IntelliGenetics name for its biotechnology business.
The market for AI held much promise in the early 1980s, spurred to a great degree by Japan’s decision in 1981 to initiate a ten-year AI initiative, which prompted Feigenbaum to write The Fifth Generation: Artificial Intelligence and Japan’s Computer Challenge to the World.Others joined him to warn about the possibility of Japan’s future dominance in the field. According to a New Republic article on the subject, “In 1983, more than a dozen American companies established the Microelectronics and Computer Technology Corporation in Austin, Texas, a consortium with a budget of about $70 million a year, including $15 million for the Alpha-Omega program, expressly devoted to countering the Japanese Fifth Generation. In 1984, the year after The Fifth Generation was published, Business Week predicted that the market for expert systems would explode from $20 million to $2.5 billion by 1993.” Despite opportunities, however, IntelliCorp shied away from investments by major corporations, opting to retain its freedom.
In short, AI was portrayed as a modern day Industrial Revolution and, as with the Internet a decade later, no one wanted to be left behind. Rather than the Industrial Revolution, however, a more apt comparison for the sudden interest in AI was perhaps the tulip hysteria that swept Holland in the 17th century, when a single bulb might equal the value of an entire ship. The AI bubble continued to grow in the mid-1980s before bursting later in the decade when the field simply failed to deliver on its much hyped promises. The Japanese effort, in particular, yielded no tangible improvements to the field. Moreover, much of the AI start-ups were run by academics rather than practical businessmen. IntelliCorp, like many of the early AI companies, relied on the LISP programming language, which in turn required a specialized computer that cost as much as $150,000 and could only be used by a single operator. The research and development units of major corporations were willing at first to invest in this technology, thus spurring initial growth in the industry, but the broader corporate marketplace failed to follow suit. It wanted AI programs that could run on its mainframes and minicomputers and did not require its people to learn LISP. As a result, AI advances began to be incorporated into general computer programs, and AI lost its status as a distinct market.
Despite making its products compatible with UNIX workstations, IntelliCorp also suffered from the diminishing interest in AI It generated $20 million in sales in 1987, but had posted six consecutive losing quarters, resulting in the laying off of 30 of its 200 workers. Feigenbaum, busy with his research, writing, and other business ventures, stepped in to serve as IntelliCorp’s chairman for a brief period in 1988 while changes in management could be sorted out. In October of that year, he relinquished the position to Thomas P. Kehler, who was then replaced as president of the company by 32-year-old Katherine C. Branscomb, a former senior vice-president of sales and marketing at Aion Corp. Her father had been the chief scientist at IBM. Rather than seeing AI technology as revolutionary, she hoped to make it evolutionary and to position IntelliCorp as a company that provided computer programmers with knowledge-based tools to help them produce better software.
IntelliCorp returned to marginal profitability, then in 1990 introduced Kappa, its first expert-system capable of running on a personal computer in a Microsoft Windows environment. Using the object-oriented tools that came out of AI research, Kappa allowed business users to portray information, whether it be statistics or design specifications, as objects that could then be manipulated. Decision-making rules were then linked to the objects and created to suit a particular business need. Far from emulating human thought, Kappa was typical of a lessambitious AI industry, which now sought to add sophistication to the basic strength of computers—the speed of completing routine operations. Kappa had been acquired the previous year from MegaKnowledge Inc. At the same time as the release of the new product, IntelliCorp acquired MegaKnowledge, paying over $3 million for the business.
Failed Merger with Knowledge Ware: 1991
Kappa, however, proved to be a disappointment. In 1991, IntelliCorp terminated a quarter of its workforce, some 50 employees, followed by a write-down on the Kappa program. When the fiscal year ended on June 30, the company reported a $14.8 million loss on sales of just $14 million. By late August the company had agreed to be acquired by Knowledge Ware Inc., which produced software engineering tools, for $34.2 million in stock. In anticipation of the deal being completed, Branscomb resigned, but Knowledge Ware posted poor quarterly results and its top executives left the company, resulting in a severe drop in the price of its stock. The merger with IntelliCorp was scuttled and Branscomb was brought back as the chief executive.
We help clients make intelligent technology decisions... and we help make technology decisions intelligent.
Intellicorp generated revenues of only $9 million in 1992, resulting in a $9 million net loss. A short time later, Branscomb resigned to become a senior vice-president of business development at Lotus Development Corporation. Chief Financial Officer Ken Haas was subsequently promoted to president of Intellicorp, and conditions began to slowly improve for the company. It gained both financing and technology from Informix, as well as a relationship with SAP, a German firm that developed resource planning software used to integrate back-office functions such as distribution, accounting, human resources, and manufacturing. SAP was interested in applying AI and object-oriented technologies into advanced versions of its products and sought out IntelliCorp for help. SAP was a fast-growing company with a large number of business customers and represented a potentially lucrative outlet for IntelliCorp’s expertise. The company deployed three core technology people to Germany, where they learned about SAP’s software and educated their SAP colleagues on AI and object-oriented programming.
Sales improved to $10.3 million in 1993 and the company’s net loss fell to $4.8 million. IntelliCorp edged closer to profitability as revenues grew to $12 million in 1994 and $17.5 million in 1995, as the net loss decreased to $3.8 million in 1994 and $2.1 million in 1995. The company’s object programming, however, still had a limited market, and much of its business was tied to just a few major customers, such as GTE Corporation and US West Communications. When GTE decided to abandon the technology, the repercussions were severe for IntelliCorp.
Revenues collapsed in 1996, falling to $11 million, while the net loss grew to $4.5 million for the year. IntelliCorp slashed its workforce by a third and once again began retooling its focus. The company’s work with SAP took center stage, as the two parties joined forces to develop products. IntelliCorp received a boost in 1996 when the accounting firm of Deloitte & Touche decided to use its software. Other major accounting firms also signed on, as did IBM. Then, in 1997, SAP chose to embed IntelliCorp technology into its business resource automation software. The emphasis on SAP would increase even further in 1998, when accounting giant Arthur Andersen elected to incorporate IntelliCorp software into its SAP technology. As a result of these developments, the outlook for IntelliCorp began to steadily improve. After generating revenues of $12.7 million in 1997 and posting a loss of $1.9 million, the company almost doubled its revenues in 1998, totaling $24.4 million. IntelliCorp still lost $700,000 for the year, due in part to the cost of expansion.
IntelliCorp incurred further costs in 1999 when it created the CRM Solutions Group to assist SAP customers in implementing CRM (customer relationship management) and eBusiness solutions. Revenues took a step back, falling to $22 million, while the company posted a $6.2 million loss. Once again, IntelliCorp was forced into a restructuring mode, as it cut staff by 20 percent and began to focus on the use of its software on the consulting side of its business rather than relying primarily on software sales. Revenues for 2000 were flat in comparison to the previous year, totaling $22.7 million, resulting in a loss of $7.1 million for the year. In March 2001, management underwent a major change. Haas was replaced as CEO by Ray Moreau, who came to IntelliCorp in August 2000 to serve as president and chief operating officer after spending the previous 12 years as an executive with Ernst & Young. Norman Wechsler, a director who owned a 31 percent stake in the company, was also named the chairman of the board. IntelliCorp posted another $7 million loss in fiscal 2001 and continued to struggle late in the calendar year as the U.S. economy officially slipped into recession. After more than 20 years in existence, IntelliCorp had managed to survive while numerous other AI companies failed, yet whether it would ever evolve into a consistently profitable business remained very much an open question.
IntelliCorp GmbH; IntelliCorp Ltd.; IntelliCorp SARL; MegaKnowledge, Inc.
Accenture; Ascential Software; Evolutionary Technologies; International Business Machines Corporation; Information Builders, Inc.; Oracle Corporation; Sun Microsystems, Inc.; Sybase, Inc.
- The company is incorporated as IntelliGenetics, Inc.
- The company goes public.
- The company name is changed to IntelliCorp, Inc.
- Katherine Branscomb is named president.
- Branscomb steps down in favor of Ken Haas.
- The company begins working with SAP.
- The CRM Solutions Group is formed.
- Haas is replaced by Ray Moreau, and Norman Wechsler is named chairman of board.
Babcock, Charles, “AI Gets Smart: Intelligent Tools,” Computerworld, July 11, 1994, p. 6.
Bozman, Jean S., “IntelliCorp Turns to Object-Oriented Mart,” Computerworld, January 18, 1993, p. 57.
Bulkeley, William, “Bright Outlook for Artificial Intelligence Yields to Slow Growth and Big Cutbacks,” Wall Street Journal, July 5, 1990. p. B1.
_____, “Stocks of Artificial Intelligence Firms Prosper, Though Some Analysts Are Advising Wariness,” Wall Street Journal, March 31, 1986, p. 1.
Johnson, Amy Helen, “Staying on Target,” Upside, June 2000, pp. 236-48.
Lewwyn, Mark, “Artificial Intelligence Firms KO’d by Reality,” USA Today, August 17, 1987, p. 3B.
Lyons, Daniel, “Artificial Intelligence Gets Real,” Forbes, November 30, 1998, pp. 176-82.
Pollack, Andrew, “Poor Decisions: Artificial Intelligence Hits Roadblocks,” San Francisco Chronicle, March 5, 1998, p. B1.
Ullman, Howard, “Machine Dreams: Future Shock for Fun and Profit,” New Republic, July 17, 1989, p. 12.
Zonana, Victor F., “Software Gang Puts ‘Experts’ in the Office,” Los Angeles Times, p. 1.