Better Data Architectures, Algorithms Take on Greater Importance

Firms seeking to extract value out of their data are looking to overcome the traditional limits of speed. The answer is better data architecture and more advanced algorithms, according to panelists at this year's FinTech Barbarians at the Gates event in New York.

Asif Alam, global business director at Thomson Reuters, said the era we are living in is transitioning from the age of information to the age of customers, as those customers become increasingly demanding.

"The customer before was looking at very simplistic structured information; now that same customer is looking for all these different kinds of variations of information, from the [traditional] sets of data available," to the data you wouldn't think to include, such as social media, Alam said.

The industry has done well to address storage needs, but now the question becomes what do you do with that data?

"How can you take structured data and unstructured data, put everything in one place and then squeeze it five million different ways?" he asked

A Need to Cut

Peter Giordano, a managing partner at Oppenheimer & Co., said players in the industry are mainly trying to cut costs related to data, which relies on reworking the firm's data architecture.

"The industry over time has become [reliant on] very large vendor platforms," he said. "What's difficult to get to are these big data sets; we have to figure out how we can extract them, turn them into quick business analytics to help us. Large, big data platforms with API integration are exactly the types of things we're executing on now," Giordano said.

For instance, he said, when combining APIs against his trading system and "mashing them together" with his customer-relationship management (CRM) data, "[I] come out with some really interesting analytics for my business folks, for my regulatory folks, but also for my clients directly," he said.

Andy Brown, CEO of Sand Hill East, pointed out that Facebook data platforms are so big now, they have superseded even the biggest banks.

Brown said it's important to invest in different algorithms, because those algos will save time in having to sift through the data. It's also necessary to utilize sophisticated blade architecture to process the data quickly.

Alam stressed that the future of data is all about customization flexibility for the customer, on top of which there will also need to be more data sharing within an organization.

"In the age of analytics, the equity desk, FX desk [and other desks] will have different kinds of data," he said. "What we need to do is find the areas of data that overlap and analyze them, in order for the data to be beneficial for all on the trading floor."

Bottom line:

* More algorithms need to be incorporated to sift through data faster.

* Companies need to work on data architecture to allow quicker analysis of data, interconnectivity and security.

* Customer demand is increasingly driving decisions behind how companies approach their data.