In 2019, Pinterest engineers built and scaled for more than 300 million people around the world, brought a more inclusive search to mobile, and made strides in shopping and computer vision by automating Shop the Look, improving Lens camera search and identifying more than 2.5 billion objects across home and fashion Pins. In addition to infrastructure and operations improvements and making mobile faster and more reliable, the team welcomed SVP of Engineering Jeremy King, introduced new home feed controls, started testing Complete the Look, and open sourced technologies.
Here are the most-read posts of the year:
- Building a real-time anomaly detection system for time series at Pinterest
- Building a Kubernetes platform at Pinterest
- The Dead Code Society
- Automating Shop the Look on Pinterest
- Optimizing Kafka for the cloud
- Developing fast & reliable iOS builds at Pinterest (Part one)
- Introducing Complete the Look: a scene-based complementary recommendation system
- Open sourcing Singer, Pinterest’s performant and reliable logging agent
- Pinterest + ktlint = ❤
- Bottoms-Up: How the Pinterest growth team decentralizes its structure
Check back with us in the new year for more stories behind Pinterest technology and teams.