by Sean McManus
Staff Writer
Businesses are always looking for new ways to stay one step ahead of the curve. Researchers aggregate the data, analysts crunch the numbers and people called knowledge managers present reports to senior management that define reasons why this product line or that market is the next big thing.
And companies, until now, have relied on generic software, sometimes in the form of simple search engine technology, to perform general sweeps of what is called structured data — the data that comes in the form of Excel spreadsheets, numerically detailing profits, expenditures, revenues and loses. It is structured data, combined with the vision of human error-prone CEOs, that guide corporate culture and strategic growth.
But what about unstructured data that arrives in the form of an e-mail from one sales representative to another stating that, hypothetically, the Southeast Asian market is about to be hit hard by a recent change in the political power base and subsequent tariff restrictions. Or a memo from the Arlington, Va. office saying that the global outlook after Sept. 11 is drastically impacting the way some clients are responding to a particular methodology. What about messages in a company chat room saying that every single person working at headquarters can’t get along with the vice president of Human Resources?
And what if all of that unstructured data could coalesce — from multiple touch points — and be presented in a way that recommends tactical, decisive action to management? In other words, what if your text mining software was proactive?
That’s what Ren Mohan was thinking when he was working as a vice president at Bank of America, where he was responsible for business analysis and design of management information systems.
“We were inundated with electronic information but were unable to use the information effectively,” said Mohan from his office inside the Technology Enterprise Center, an incubator on the Southside. “All of the existing software offered only a partial solution, such as searching for one document at a time. It was hard to form comprehensive solutions.”
So Mohan and his wife Usha, a former business consultant with Deloitte & Touche, wrote an algorithm that would categorize, analyze and share the vast pool of knowledge assets contained in unstructured information.
They started a company called Intelligenxia. Ren Mohan is the CEO and Usha Mohan is president. Together, they have developed proprietary technology that will automatically index and extract “concepts” from unstructured data like e-mails, faxes, websites, web pages, magazine articles, scanned documents, PDF files and spreadsheets based on the meaning of the text contained in the data, allowing users to search based on intent versus keywords.
Then, using principles of natural language, statistical analysis and probabilistic theory, the data is categorized and put into folders that can be accessed by knowledge managers to make intelligent recommendations for the future.
“It’s kind of like Cliff’s Notes for unstructured data,” said Chris Demetree, who recently joined Intelligenxia. “So much information is stored inside a company’s Intranet but nobody knows how to access it. This brings it to the surface in a usable way.”
What Intelligenxia software will do is resurrect old, depreciative data and put it to practical use by analyzing the textual information, enabling companies to capture information trends and perform proactive analysis.
“There is tremendous potential for applications across industry,” said Demetree. “From law firms to health care to government.”
For example, say a doctor is interested in finding out about rising health risks among a specific population. The old way was to hunt through pages and pages of patient histories and try to locate similarities across the country or the world. Intelligenxia does that, running a global search of health records and producing a report of which diseases are becoming more pervasive for certain age groups in certain geographies.
“It’s a natural convergence process,” said Mohan. “It will eliminate hours of wasted time spent on painstaking research and at the same time make all research methods highly efficient.”
“If Sept. 11 taught us anything its that there’s more to data collecting than just storing it,” said Demetree. “It’s only useful when you analyze it and form productive conclusions.”
One of Intelligenxia’s primary selling points is that its utility flows in conjunction with the rising corporate need for data storage. While companies like EMC and Compaq vie to sell large boxes to store vast quantities of electronic information, Mohan and the Intelligenxia team are right there to say, “Now let’s make that data valuable.” Intelligenxia hopes to usher in a wave of text warehousing.
“Law firms will be able to conduct targeted searches for specific case law, the Defense Department can search entire networks for certain individuals for character analysis and reports can be generated identifying potential opportunities that nobody even thought of,” said Mohan. “Big companies will be able to identify best practices across multiple processes and through multiple departments.”
In the new start-up climate, Intelligenxia is working to generate funding and hire top talent. They’ve got about a million dollars and a competitive analysis firm in the middle of a beta test. They are also working with a law firm and a medical firm. Intelligenxia has patented the original algorithm and ones with more direct vertical applications are pending.