Prescient AI
Ask about your own numbers and get an answer grounded in your Base and Halo model. It works alongside your CSM and never invents a number.
Prescient AI is an analyst built into the platform. It reads the same Base and Halo model behind your Home and Performance views, the same spend, modeled ROAS, cost per new customer, and halo effects, and answers questions about it in plain language. Ask it what you would ask an analyst, and get the answer without stitching together five tabs to find it.
Ask in plain language
Type a question the way you would ask a teammate. It reads your account, answers from your own model, and points you to the next thing worth checking. It works off the same view behind Home and Performance, so what it tells you matches what is already on your dashboards.
Recaps
Ask for a recap of any period or channel. "Recap the last 30 days by channel" comes back with spend, modeled ROAS, and halo effects straight from your model.
Diagnostics
When two numbers disagree, ask why. "Why does my modeled ROAS differ from what Meta reports" walks you through where the platform-reported and modeled figures split, using your own data.
Comparisons
Compare two periods, two channels, or your modeled numbers against what the platforms report, without stopping to build a report.
Next best action
Ask what deserves attention and it surfaces a next step from your model. You show up to the budget review with the context already gathered, and you decide whether to act on it.
Built with guardrails
It works alongside your CSM, who stays your partner for the bigger strategic calls. Answers stay inside your own data. It will not invent a number, and it will not fall back on an outside benchmark.
Part of Prescient Horizons.