Insights from Building Enterprise-Grade ML Experiences at Asana

In the past year, Asana embarked on building its first user-facing machine learning (ML-powered) feature using data and AI. We will share key learnings from building our first ML-powered features and having it run for millions of users every day in production. The talk will include: - Modeling/embedding different Asana Work Graph (R) Objects to surface the most relevant existing and new work content to users - Monitoring model performance offline while training and online during production through A/B testing - Building and scaling batch and real-time ML platform to support a variety of ML-powered use cases for thousands of customers - Using Databricks as a data and computation foundation to power data science solutions to enable data scientists to easily develop and deploy ML model This talk will specifically be beneficial for audience members who are aiming to build or enhance ML capabilities in their organizations by providing key learnings and insights from Asana's journey.