[DOCS-2738] AI Optimizer

Imagine you are considering two playlist strategies to drive revenue. One is curated by your editorial team and another is powered by automated recommendations. How can you dynamically measure which performs better over time as viewer behavior and advertiser demand change? You could create a traffic split in your decision tree and manually monitor results, or you could use an AI optimizer to dynamically automate this process.

An AI optimizer node is an intelligent strategy component that evaluates branch performance and automatically adjusts traffic distribution. It continuously learns from real-time revenue data to determine which experience delivers the highest value.

Key Benefits

• Evaluates and rebalances traffic distribution daily
• Adapts to shifts in audience engagement and advertiser performance
• Uses monetization data from the ad server to guide optimization decisions

The AI optimizer starts with an equal traffic split across branches, then continuously reallocates traffic toward the best-performing branch.

Prerequisite

Add an AI optimizer node

Follow these steps to create an AI optimizer node:

1. In your decision tree, click > AI optimizer branch. The Create AI optimizer panel opens.
2. Enter an AI optimizer name.
3. Enter the Branch 1 name.
4. Enter the Branch 2 name.
5. (Optional) Name an additional branch:
   1. Click + Add Branch. A new field Branch X name field appears.
   2. Enter a branch name
   3. Repeat steps 5a-b to name another branch.

      📘

      You can add up to 13 additional branches (a total of 15 branches).

6. Click Apply. The strategy tree appears with the AI optimizer node and branches.
7. For each branch, define the experience with one or more of the following:
   1. Condition branch
   2. Traffic split branch
   3. JW video experience
8. Click Save.

You can monitor the node performance in your strategy analytics.