See MapD and the rest of the FinDEVr New York speaker lineup on March 21 & 22. Register and save 10% with the promo code FinDEVrBlog.

Interview with Todd Mostak, MapD’s CEO & Founder:

Where did you start your career and how did you gain the experience needed to run the tech side of your company?

I think most people would consider my path to becoming a CEO of a startup to be unorthodox – even for startups.

I graduated from the University of North Carolina at Chapel Hill with a degree in Economics and Anthropology. After graduation I moved to Syria to teach English for Berlitz while learning Arabic myself. I found the language and culture fascinating and would go on to study Arabic at the American University in Cairo where I also worked as a translator and occasional writer for the Egyptian newspaper Al-Masry Al-Youm.

With my eye on academia I went to Harvard, where I started to work on my master’s degree in Middle Eastern Studies with a focus on how social media impacted the Arab Spring. My thesis involved performing analytics on hundreds of millions of tweets – and I didn’t have access to the computational horsepower or visual analytics tools to do it effectively. I had cross-registered for a database systems class at MIT where I built a prototype database that ran on graphics cards rather than on conventional processors. This allowed me to do and see things in the tweets that I would have never found using traditional methods.

With the encouragement of MIT Professors Sam Madden and Michael Stonebraker I ended up joining their Computer Science and Artificial Intelligence Lab (CSAIL) and working full time on this concept of a fast system (database + datavis). In 2013, we won Nvidia’s Early Stage Challenge and decided to make MapD our sole focus.

From a technologist’s perspective, what’s unique and game-changing about your technology?

MapD has leveraged the parallel processing power of GPU to create the world’s fastest data exploration engine. By pairing a lightning fast, SQL-compliant, GPU database with a cutting edge visual analytics engine, MapD allows multi-billion record datasets to be queried and visualized in milliseconds. For organizations, this means that they can tackle problems that are far larger, far more complex, far more difficult than previously imagined – without losing any grain level detail.

Speed of thought data exploration used to stop at tens of millions of rows – constrained by both the performance of CPU-bound databases, but also by the performance of in-browser visualization solutions. To go beyond those limitations meant waiting minutes, if not hours to see your results. MapD’s GPU-powered software reframes that conversation.

Now hundreds of billions of rows can be scanned and visualized in less than a second. The ability to truly explore massive datasets, the way humans crave, rapidly testing and validating hypotheses with millisecond response times is now possible. Further, the ability to visualize that data – at any level desired, comes into frame; from a model of every wind pattern in the Pacific to a single gust of wind, from 1.2 billion taxi rides in NYC to a single trip to JFK, from hundreds of millions of political donations to those on your street.

The biggest challenges facing the financial services industry invariably involve data. Creating a framework to evaluate every datapoint, without downsampling or indexing means we can engage broadly, rapidly and successfully on issues like alpha generation, fraud, AML and operational risk.

Tell us about your favorite implementation of your solution/technology.

I would be lying if I didn’t say that my favorite implementation was our Tweetmap. It was the first thing I built and was the primary tool to explain the power of speed at scale for several years. We have since expanded, not just in terms of other public demos but in terms of what our clients are doing.

This is what excites me the most – the types of problems our clients are applying our technology to solve. In financial services, it is clearly our hedge fund clients, who are looking at billions of rows of data (production data, clickstream data, product data, weather data) to find nuggets of alpha. These are challenges that were simply too big for their existing infrastructure and their ability to look across all of it means that they can “see” more than their competitors. They like that.

FinDEVr New York 2017 is partnered with American Banker, BayPay Forum, BiometricUpdate.com, Breaking Banks, Byte Academy, Canadian Trade Commissioner Service, Celent, CIOReview, Cointelegraph, Colloquy, Distributed, Economic Journal, Empire Startups, Femtech Leaders, Finmaps, Fintech Finance, Harrington Starr, Level39, Mercator Advisory Group, The Paypers, SecuritySolutionsWatch.com, Swiss Finance + Technology Association, and Women Who Code.