The panelists, Richard Daley, Founder and Chief Strategy Officer of Pentaho; Edouard Sevan-Schreiber, Director for Solution Architecture of 10gen; Kathleen Rohrecker, Director of Marketing of Revolution Analytics; Patrick Angeles, Solution Architect of Cloudera, were questioned by the moderator.
Andrew Brust asked the panelists if there ever would be an intersection between big data and NoSQL. Richard Daley responded that he started with commoditizing BI (business intelligence). “Big data needs to coexist, consolidate data together,” he said. “Monetize that value.” He went on to say that one “can’t just take a BI product and attach something to it…. It’s going to be a coexistence.”
Eduard Sevan-Schreiber responded that the traditional model of data is no longer sufficient in this modern age. “Big data moves so fast, that it can’t keep up,” he said. Changing a data model takes months, so MondoDB is one of the solutions.
Kathleen Rohrecker said that everyone uses “analytics” because it is a hot word right now. She underlined that “you need to make a distinction between what type of analysis your company needs—reporting, analyzing, predicting, and so on.” She basically summed up analytics as “advanced statistics” and went on to say that all data is “complementary.”
The moderator intervened at this moment to ask a question directed at Patrick Angeles. He asked Angeles about a refinery model, if a dupe is used for unstructured data, cleaned up and sent. “Is it an assembly line thing?” Brust asked. Angeles responded, “No, it’s going to be more fluid than that.” He emphasized that “[everyone is] all working together to enhance our abilities.” He also added that “data is worth something when it is associated with other things.”
Brust then asked Rohrecker if “we are going to get past unstructured data. Will there be predictive analytics to the user and when?” Rohrecker responded that there is “always a tradeoff” and that “it is an iterative process.” The multi-step process consists of analyzing, hypothesizing, ensemble, and assembling data.
The panelists, summing up, said that non-Fortune 500 companies were the ones taking more risks than the established companies. They’re the ones that implement the data even if it is “a bit quirky.” It was also revealed that small- to medium-size companies use large amounts of data—more than big companies—and they all use tech analytics.
The event was wrapped up by reiterating the panelists’ intentions: providing open-source alternatives to commercial products.