Machine Learning at Discord

Gaurav Chakravorty on building industrial grade machine learning systems for search, recommenders, and personalization systems.

SubscribeApple • Android • Spotify • Stitcher • Google • AntennaPod • RSS.

Gaurav Chakravorty, is a Senior Manager at  Discord, where he leads the team responsible for machine learning models for search and notification. Prior to discord Gaurav was a manager at Google where he led the team responsible for personalized podcast recommendations.

Gaurav Chakravorty:

There are two broad areas of applied machine learning within Discord: (1) anti abuse and safety,  and (2) applied ML for discovery notification and discovery. Supporting both of these is the ML platform effort. Along with the ML effort is the golden data, the data engineering, data infrastructure, and certainly the data science initiatives. Now, within discovery and notifications, there are problems like, recommendations, ranking, semantic search, notifications, and app search. This also includes content summarization, which involves extracting multi-tenant knowledge graphs of interests.

Highlights in the video version:

Related content:

Download the FREE report below:

[Image: Cartography from Pixabay]