AdMagic was not built by two people who read about the ROAS problem in a case study. It was built by two people who lived inside it: one who spent years negotiating and defending media budgets, and one who builds the AI agents now trying to fix what that budget kept getting wrong.

The marketer: $4M+ managed, $2.1M+ generated, and a lot of uncomfortable meetings

Before AdMagic, one of our founders spent years running paid social for other people’s businesses through prior agency work: $4M plus in ad spend managed, $2.1M plus in client revenue generated. Numbers like that sound like a highlight reel. They are not. They are years of sitting across from a founder or a CMO, pointing at a dashboard that says 4x ROAS, and still having to explain why the bank balance does not agree with the dashboard. That gap, between what the ad platform reports and what the business actually keeps, is the entire reason AdMagic exists. It is not a theory someone picked up at a conference. It is a problem our own team has had to explain to a client’s face, more than once, with no good tool to point to.

The engineer: building the agent, not just the dashboard

The other half of the founding team builds AI agents for a living, and that background shows up in how AdMagic is actually put together. AdMagic’s Context Engine ties every dollar of ad spend and every click to true net profit, not just ROAS. TRIBE v2, AdMagic’s neuro-creative scoring system, is built on fMRI and EEG data from 750 participants and is used to score a creative before a single dollar goes out the door. Layered on top is an AI CFO and CMO agent that proposes recommendations with a reason and an audit trail attached to each one. It does not act on its own. A human approves before anything executes. That last part was a deliberate call, not a limitation. A marketer who has had to defend a budget to finance does not want a black box making decisions for her. She wants a second set of eyes that shows its work.

Why the pairing matters

Plenty of ad-tech tools get built by engineers who have never had to justify a media budget to a CFO, or by marketers who have never had to specify what an agent should and should not be allowed to do on its own. AdMagic is being built by both, together, inside Toronto Metropolitan University’s Innovation Incubators and Momentum-SDZ program, one person insisting the numbers mean something real to the business, the other insisting the software actually deserves to be trusted. AdMagic is still in active MVP development. We are not claiming live customers or a public beta today. We are building toward our first design partners, with a target of roughly 10, around January 2027. What we do have is a founding team that has already spent years on the exact problem AdMagic is meant to solve, and a product being built by someone whose job is literally to make AI agents worth trusting with a budget. Want in before the general rollout? Join the AdMagic waitlist and get early access as we get closer to launch.