Build Stuff'16 Ukraine

Have you ever wondered where your favorite feature came from? Was it influenced by a feature in another language? How are the different programming languages even related? I spent a couple months researching the history of some programming languages, and wanted to share that with you. In this talk, I cover the history of the ML family from approximately the dawn of time, eventually focusing on F# specifically.

Functional programming has been a hype for few years already. Almost every developer should already know that it isn’t just for scientific or financial programming but can be used for business applications as well. However it has not got wide adoption yet. It is still considered as a something like rocket science by many developers.
In this talk Serhiy will show how functional approach can be used in development of real life application. We will cover the process from development to deployment and end-to-end testing. And we will use F# everywhere.
Come to my talk and you’ll see how simple and elegant your code may be if you choose functional programming.

Want to simplify state management, improve scalability and reduce the code base? Rewrite your system in F# and use the actor model (Akka.NET). The functional language discourages use of mutable state and actors contribute to efficient communication, routing and scalability. And what about the code base? The code metrics speak for themselves. This talk is about real-world project and summarizes experience writing a robust and performant message based system for distributing media files to the cloud.

Orleans is a runtime and programming model for building distributed high-scale systems based on actor model. Orleans is an implementation of an improved actor model that borrows heavily from Erlang, adds static typing, message indirection and actor virtualization, exposing them in an integrated programming model known as "Virtual Actors". Orleans has been used to build multiple production services, including back-end services for a number of popular games, like Halo 4. Unfortunately, despite being truly unique and practical project, Orleans didn't make a lot of fans in F# community, due to its dependency on static code generation, verbosity of programming model and complex configuration. Orleankka - is a complementary API for Orleans built with functional paradigm in mind. It brings proper support for F#, custom DSL that will be familiar to all Akka users, and heaps of other usability improvements over the original API. Be aware: demo heavy talk! Lots of code examples demonstrating major Orleans features done with Orleankka’s API will be presented.

Just like traditional applications development, machine learning involves writing code. One aspect where the two differ is the workflow. While software development follows a fairly linear process (design, develop, and deploy a feature), machine learning is a different beast. You work on a single feature, which is never 100% complete. You constantly run experiments, and re-design your model in depth at a rapid pace. Traditional tests are entirely useless. Validating whether you are on the right track takes minutes, if not hours.
In this talk, we will take the example of a Machine Learning competition we recently participated in, the Kaggle Home Depot competition, to illustrate what "doing Machine Learning" looks like. We will explain the challenges we faced, and how we tackled them, setting up a harness to easily create and run experiments, while keeping our sanity. We will also draw comparisons with traditional software development, and highlight how some ideas translate from one context to the other, adapted to different constraints.