Cortex is a group of researchers, engineers and scientists working together to build state of the art machine learning technologies within Twitter.
Twitter is what’s happening in the world right now and operating at such a global scale brings massive research and engineering challenges. Cortex exists to solve these challenges through development and application of new machine learning techniques, aiming to ensure that users have the best possible experience across all of Twitter’s platforms and media - from connecting users with the best content to enhancing video quality.
Our world-class team of engineers and researchers work across all levels of the Twitter stack and across all products, with core technologies developed by Cortex becoming essential parts of Twitter, Vine and Periscope on both Web and Mobile. We’re committed to open source and open publishing and work on long-term research alongside fast product development.
Latest Updates
-
The blog details how Cortex is working toward developing models, related tooling, and infrastructure with the objective of making embeddings a first-class citizen within Twitter's ML platform… Read more
-
Cortex automates machine learning pipelines at Twitter with Airflow.… Read more
-
Twitter Cortex migrates deep learning framework from Lua Torch to TensorFlow.… Read more
-
Today we are excited to announce torch-decisiontree: a python machine learning library using gradient boosted decision trees and random forests for sparse and dense tens… Read more