Panel: Big Data Can Be ‘Democratized’ For The Masses
Big data is a big deal, and it’s not too late for people who feel they know little or nothing about the collection, organization, storage, and retrieval of massive amounts of information to start paying attention. The trend isn’t going anywhere.
Those were some of the themes of a panel discussion held on Monday in Madison, WI, that involved leaders at Wisconsin-based tech companies.
The discussion, which was put on by the Greater Madison Chamber of Commerce and Big Data Madison Meetup, touched on a number of topics, including approaches for giving employees the needed access and tools to fetch and analyze data, and some of the challenges and opportunities that arise with combining the information a company collects itself with outside data sources.
“Everything we do is data-driven,” says panelist Andy Jennings, user acquisition coordinator at PerBlue, a Madison-based startup that develops games for mobile devices. “I think PerBlue has done a good job at making it part of the culture,’’ he says. The mobile games industry is maybe 10 years old, and it evolves with every new device that’s introduced, adds Jennings. “And the advertising gets more and more sophisticated, so you have to keep track of that.”
Putting data at the heart of how a business thinks and operates likely comes easier at a smaller company like PerBlue than it does at one with hundreds of workers, such as Wisconsin Rapids, WI-based Renaissance Learning. Bipin Karunakaran is a senior data scientist at Renaissance, which says its cloud-based educational software is used in more than 60 countries across the globe.
Karunakaran, who was part of Monday’s panel, says that executives and number-crunching employees can talk about analytics and big data all they want; but weaving data-driven decision-making into the fabric of an organization involves getting large numbers of workers to understand why the information is important, and how they can get their hands on it.
One of the current initiatives at Renaissance makes use of Redshift, a hosted data warehouse product that’s part of Amazon Web Services. Karunakaran calls the technology “really efficient.”
“The objective is really to democratize data so that more people in the organization can use it,” he says.
People need to be able to walk before they can run, though, says Greg Tracy. He was the third member of Monday’s panel, and is co-founder and chief technology officer of Propeller Health, a Madison-based startup that creates software and hardware to help patients with asthma and chronic obstructive pulmonary disease.
He says that from the beginning—Propeller was launched in 2010, originally under the name Asthmapolis—his company “stored anything and everything we could possibly think of as we built stuff.”
Despite its inveterate emphasis on data collection, Propeller found that not all of its employees were comfortable navigating the information when given the ability to pore over it.
“They were like, ‘Oh my gosh, I’m overwhelmed by all of this crap. Could you please build some tooling to make it actually useable?,’” he says.
Propeller then made what Tracy calls a “non-trivial investment” in building its own framework to gather data from various sources. Today, when employees want to create new databases or information sets, they are almost without exception required to conform to this framework.
“That’s something that I’m really happy about in hindsight, that we made that investment,” he says.
Aside from Propeller’s own information gathering, another source of data is governments and public agencies, both at the local and federal levels.
Tracy says Propeller has worked with the municipal government of Louisville, KY, to create a program that provides some asthmatic residents there with free smart sensors for their inhalers. The goal for Propeller and city leaders is to gain insight into when and where people with asthma experience symptoms.
He says Propeller is also able to combine the information its sensors and inhalers gather with data on air quality and weather from the Environmental Protection Agency and the National Oceanic and Atmospheric Administration.
“We’re putting lots of government data sources to work inside of our system,” he says.
Tracy says the biggest challenge when it comes to integrating government data is what he sees as a lack of technical standards. For example, he says there are groups all over the country collecting pollen data, but they’re not always consistent in how they document and publish the information.
One leader in Madison who bridges the worlds of government and technology is Maurice Cheeks. He serves on the Madison City Council, and is also vice president of business development at MIOsoft, a software company based in town that helps customers tackle big data problems, including improving and validating the quality of the information they collect.
Cheeks, who was not part of Monday’s panel discussion, says that most organizations today, even small ones, have more data at their disposal than they know what to do with. But even if leaders at these groups know what types of data are being stored, and how to get at the information, many of them are still a ways off from using their own data to its full potential.
“Data, I think historically has been something that lives in this little siloed area of an organization, whether that’s IT or finance or maybe marketing,” he says. “One of the things that’s changing about the world is that companies are starting to recognize that their data is one of their greatest assets.”