Real-time machine learning is increasingly important, driven largely by two factors: (1) the imminent death of the 3rd-party cookie, and (2) increased recognition around the use of contextual information (which often reflects the user’s current state of mind) when making decisions around how to engage with a user. Real-time machine learning can be used to increase paywall conversions, choose the right pop-up messaging, enable dynamic decisioning across multiple offers, and implement next-best-offer/action experiences.

All Cortex accounts contain a new set of functionality which makes it easier to build and deploy real-time inference using Cortex. Below are a couple examples of the changes you will notice within your Cortex account.

First, when building out a new machine learning pipeline you can now toggle between Batch and Real-Time pipelines. Note that both Future Events and prescriptive Uplift Pipelines are available in a real-time context. Combining real-time machine learning with the Cortex Decisioning SDK enables a business to leverage both batch and real-time user data for experiences like dynamic decisioning.

You’ll also note that when building a real-time pipeline you have the option of selecting conversion events happening within seconds of an intervention. An intervention could be any number of actions on a user including a shown offer or a paywall landing page.

In the example above, a Cortex user is building a real-time pipeline to predict who will convert within 15 seconds of seeing an onsite offer.

There are numerous additional changes in your Cortex account associated with real-time machine learning. As you explore these features, please don’t hesitate to reach out to!

Want to Learn More?

Schedule a demo and talk to a product specialist about how Vidora’s machine learning pipelines can speed up your ML deployment and ultimately save you money.