San Francisco: Crunching Gigabytes Locally

Notice that this "third Wednesday" meetup is at Mesosphere, not Symantec. And notice that Mesosphere has a new address in the Financial District! It is now a couple blocks NORTH, not east, of the Montgomery Street BART.

As C++ fans, we know well how effective the language can be in the hands of an experienced engineer. In the talk this claim is backed up by matching various C and C++ solutions against each other, and against other ways to crunch through data, including Python/Cython and SQL- and NoSQL-based data stores. After the comparisons, the trick to dynamically compiling and loading C++ code is introduced and evaluated as well. Spoiler alert: It is lightning fast and overall awesome.

Dima Korolev is passionate about data, machine learning, infrastructure, reliability, high performance, and C++. In his previous life he used to be a competitive programmer, and his background includes coaching and teaching, both of CS fundamentals and of data mining fundamentals. If the words "single source of truth," "live schema evolution," "zero-downtime master flips," or "analytical differentiation" excite you, you and Dima will likely find plenty of mutual interests.