The company Narrative Sciences unique approach to addressing the Big Data Challenge, as the dilemma of managing fast-growing volumes of information is sometimes referred to, comes from the firms unusual genesis. Kris Hammond, the founder of Intelligent Information Laboratory, Northwestern Universitys artificial intelligence lab (who later founded Narrative Science in 2009 along with Larry Birnbaum, another Northwestern computer science professor, and Stuart Frankel, previously an executive with digital advertising company DoubleClick), decided early on to partner with Northwesterns respected journalism school to test and refine his labs capabilities.
A collaboration between engineers and journalism students ensued, allowing the two groups to apply advances in artificial intelligence to content and news generation. In time, Hammond and his partners realized that the technology was also relevant to data-driven businesses like finance.
To date, the Chicago-based start-up has raised $7.5 million in two rounds of venture capital funding from Battery Ventures, SV Angel and Northwestern University. Not surprisingly, Roger Lee, a general partner of Battery Ventures, Narrative Sciences lead venture capital investor, is optimistic about the firms outlook for growth among Wall Street firms. I see the intersection of two trends the explosion of data on Wall Street and the fact that the Big Data trend will only continue to grow and accelerate as well as the need to manage it, as presenting a good opportunity for this technology, says Lee.
He projects that three types of users could benefit from use of the Quill, Narrative Sciences new AI-based software platform aimed at helping individuals in high data-generating industries to better analyze and comprehend data more quickly. The first type of potential users are those involved in the creation of external communications, such as CEOs and CFOs, along with portfolio managers and investment research teams. (I see the Schwabs and Fidelitys of the world making good use of this type of technology for their customers, Lee says.) The second type includes individuals involved in internal decision making, including top management, risk managers and operations support personnel. And the third type are individuals involved in ensuring that their firm is compliant with all the new regulations generated by the SEC and FINRA.
Others are skeptical about the expanded use and possible overuse of AI technology. Tom Garske, an associate partner at Citihub, a consulting firm that targets the financial services industry, says that although he recognizes the potential benefits of expanded use of artificial intelligence at financial firms, overuse of analysis can lead to even greater levels of information to review and a form of paralysis by analysis.
How do you comb through all this information, even in narrative form, so that you are actually providing results? And you have to be careful about how all this raw data is interpreted, Garske says, pointing out that during a time of constrained budgets at financial firms, most will take a wait-and-see posture before trying such technology.
Paul Rowady, a senior analyst at research firm Tabb Group, says that although he is a big fan of applying pattern-recognition technology and AI capabilities to Wall Streets vast data stores, such efforts will have to contend with the low signal-to-noise ratio, or the fact that it is very difficult to extract reliable decision-support signals from all the structured data that is out there.
Rowady also believes firmly that data supplied in a visual format is best for people who have to make decisions quickly. There is a low density of information in the narrative format; it is much more effective to transfer important information via pictures, Rowady says. Taking information from a dense format and converting it into a less dense format is heading in the wrong direction.
Hammond of Narrative Science, however, thinks there is room for more than one way for data to be served up by Wall Street firms and their clients. For many people, raw data is often indecipherable, and although there are many people at financial firms who regularly review and assess visual data, the mounting volumes are increasingly unmanageable. They too can benefit from our offering, he says. Others simply need assistance in understanding the gold in all that data.