The enterprise Machine Learning (ML) landscape is increasingly trending toward real-time inference. This is especially true when it comes to powering dynamic user journeys onsite. Done right, Real-Time ML can be used to intelligently determine which customers should receive which experiences in order to maximize revenue, conversions, or engagement. There’s no more powerful tool for a digital product team.
In a recent article on Towards Data Science, Vidora product manager Michael Firn discusses various approaches to Real-Time ML, and makes a case for product teams to adopt an implementation approach when building dynamic user journeys. More specifically, the article examines topics such as:
- What is Real-Time ML?
- What are the advantages of Real-Time ML (compared to Batch) for dynamic user journeys?
- What implementation challenges does Real-Time ML pose, specifically around feature engineering?
- How B2C organizations can navigate these challenges through a best-of-both worlds approach that leverages both in-session and historical features for real-time inference
- An example related to dynamic paywalls for a subscription media organization
Check out the full article for more details and let us know if you’re interested in learning more!
