The situation
A regional sports franchise had a sharp, distinctive brand voice. Local, irreverent, written for fans who knew the league inside out. They wanted to extend that voice across multiple AI touchpoints: a content-creation agent, an audience-targeting agent, and an evaluation agent. Each touchpoint had different requirements. The content agent needed to sound like the brand. The audience agent needed to sound like a specific fan persona. The evaluation agent needed to stay neutral and not adopt any voice at all.
What we found
The agents needed a shared method but different instructions for how to speak and evaluate material. Combining all the voice rules would make it difficult to keep those roles distinct, especially for the evaluation agent.
What we built
A layered instruction architecture. At the base, the shared methodology that all agents respect: the brand's research-driven approach to audience response. Above that, per-agent voice and persona instructions that override or extend the base for that agent specifically. The evaluation agent runs on the base only, with explicit instructions to stay out of persona. Brand-voice agents pull from a separate sub-stack that gets swapped in without touching the foundation.
The separation also made it possible to add another voice through configuration while preserving the shared method.
What changed
Each agent sounded like itself, not a homogenized version of the brand. The content agent produced output the brand team approved without revision. The audience agent could be re-tuned to a different persona without rewriting the foundation. The evaluation agent didn't accidentally rate persona-laden output more highly because it shared no voice rules with the agents producing it.