Fangjin Yang, creator of Druid, shows how approximation algorithms can help system scale out linearly and process huge amount of data quickly with small memory footprint.
Yoni Goldberg describes some of the technological innovations that have helped Gilt to reach its current size, and highlight some of the core challenges that the company's engineering team faces.
Jon Hoffman discusses the general architecture, storage systems and development practices created to handle the ever increasing volume and complexity at Foursquare.
Wesley Chow presents Chartbeat's real-time analytics platform and how able to handle the requests in a cost efficient manner using a custom written analytics engine in C and Lua.
Aish Fenton discusses Netflix' machine learning algorithms, including distributed Neural Networks on AWS GPUs, providing insight into offline experimentation and online AB testing.
Avleen Vig discusses the changes Etsy has implemented to scale continuous deployments over the last 12 months, in both software and infrastructure.
Fred Hebert introduces two strategies for handling overload -load-shedding and back-pressure- along with different ways to make them work in Erlang focusing on the importance of planning for overload.
Benoît Chesneau discusses creating, scaling and reusing HTTP connections, summarizing techniques used to reduce memory usage in Erlang and ways to handle massive client connections efficiently.
Tim Fox discusses the design principles and motivation behind Vert.x and why the future is reactive.
Aviran Mordo introduces Wix's architecture, a highly available eventually consistent system, along with patterns for rendering many websites with a relatively small number of servers.
Brian Troutwine shares insight on using Erlang for a highly concurrent and very low latency bidding system implemented by Adroll.
Michael Dowden introduces JMeter and explains how to develop a data-driven methodology to determine some of the limits of a web application: max number of concurrent users, bottlenecks, etc.