One of the best ways to drive incremental revenues with machine learning is to leverage real-time user behaviors to drive conversions through experiences like next-best-action and next-best-offer (here’s more info on Vidora’s approach to enabling real-time decisioning). Real-time data enables one to target based on in-session activity and opens up the opportunity to target first-time users, anonymous users, and existing user intent.
We’ve seen real-time user data increase next-best-action conversions by 40%+ when compared to only using batch data.
One of our goals with Cortex is to make it fast and easy for our customers to integrate real-time machine learning decisioning. With Twilio Segment, sending real-time data to the Vidora Decisioning SDK is relatively straightforward, only requiring a small change to the Segment SDK. Decisions are then available from the Decisioning SDK which can be used to increase conversions, next-best-action, and next-best-offer experiences.
In summary, the steps to enable real-time machine learning decisioning using Vidora Cortex and Twilio Segment are –
- Configure the Segment SDK to send data directly to Cortex
- Build real-time machine learning pipelines in Cortex (here’s a video on how to build real-time machine learning pipelines) – this is done through the Cortex UI
- Act on the Cortex decisions for next-best-action, etc.
This integration will make it faster and easier for joint Twilio Segment and Vidora customers to enable advanced machine learning experiences like next-best-action and dynamic decisioning.
Let us know if you have questions!
