Identify User Journeys at Pinterest Lin Zhu | Sr. Staff Machine Learning Engineer Jaewon Yang | Principal Machine Learning Engineer Ravi Kiran Holur Vijay | Director, Machine Learning …

Pint (Python In The News) curates the latest news, updates, and trends in the Python programming language ecosystem. Developers can stay informed about new Python releases, library updates, and community events, as well as discover Python-related articles and resources from around the web. Additionally, Pint highlights Python use cases in various domains such as web development, data science, and machine learning, offering insights into Python's versatility and widespread adoption.

Pinterest Engineering

Pinterest developed a user journey identification system to understand long-term user goals beyond immediate interests. The system uses dynamic keyword extraction and hierarchical clustering to identify journey patterns from search history, engagement data, and board content. Key components include journey naming using ranking models and LLMs, a point-wise ranking model with diversification, and stage prediction to classify journeys as situational or evergreen. Initial deployment in journey-aware notifications showed 88% higher email click rates and 32% higher push open rates compared to interest-based notifications. The team is now consolidating components using fine-tuned LLMs (Qwen) with Ray batch inference for improved scalability.

Identify User Journeys at Pinterest

Get Pinterest Engineering’s stories in your inbox

1. User Journey Extraction and Clustering