Swift is a strongly-typed programming language that is compiled to native executables for ultimate performance. Until now, most of the Deep Learning was done in Python. A vast majority of data scientists also work in Python. At the same time, software engineering moves to Functional Programming while evolving various techniques, such as test/behavior-driven development, property-based testing, continuous delivery and so on that is much better enabled by the type-based approaches. We hope that S4TF, supporting TensorFlow at compiler level, with its Apple and Google origins, will herald a revolution in the full stack Machine Learning deployment from the server to the edge.
We need a downtown SF location for this meetup providing food and beer for developers!
Eugene Burmako was a student of Martin Odersky, the creator of Scala at EPFL, and the creator of Macro Paradise. After that he was a team lead at Twitter, developing Reasonable Scala (rsc).
Recently Eugene joined Google where he works with Chris Lattner, the creator of Swift, on Swift for TensorFlow. As a compiler specialist working on developer effectiveness, Eugene has unique insights into the ecosystems of both languages. While his first impressions are still fresh, Eugene will share his impressions of Swift with us.