Insights Investing in privacy enhancing tech to advance transparency in ML

We’re investing in privacy-enhancing technologies (PET) to pioneer new methods of public accountability and access to data in a manner that respects and protects the privacy of people on Twitter.

Insights Model-based candidate generation for account recommendations

In this blog post, we share how we improved account recommendations with a model-based candidate generation framework.

Insights A hybrid approach to personalize notification volume

In this blog, we’ll share how we use machine learning techniques to ensure people see the right amount of relevant and timely notifications.

Infrastructure Advancing Jupyter Notebooks at Twitter - Part 1

Twitter Notebook is Twitter’s internal notebook solution. It provides a first-class development environment to Data Scientist and Machine Learning practitioners at Twitter. This blog shares features o

Infrastructure How we built Twitter’s highly reliable ads pacing service

In this blog, we describe how we separate Twitter’s pacing system from the serving stack to an independent service.

Insights GNNs through the lens of differential geometry and algebraic topology

In this post, we show how tools from the fields of differential geometry and algebraic topology can be used to reinterpret GNNs and address some of their common plights in a principled way.

Insights Speeding up Transformer CPU inference in Google Cloud

This blog post shares optimization findings to speed up Transformer-based models’ CPU inference and improve computational demand in Google Cloud.

Insights Forecasting SQL query resource usage with machine learning

How we apply machine learning techniques at Twitter to forecast SQL query resource utilization during the development and maintenance of our large-scale SQL system.

Insights How we rolled out security keys at Twitter

In this blog, we share how we’ve accelerated efforts to increase the use of security keys to prevent phishing attacks at Twitter.

Infrastructure Processing billions of events in real time at Twitter

How we process large scale data in real time at Twitter while reducing latency and improving accuracy.

See more