Keep raw prompts out of partner ad requests.
The data firewall checks and reduces the context used for ad matching before a request reaches an advertising partner. Raw prompts stay out of that partner request.
The delivery path can use broad categories such as travel, education, coding, or finance, plus locale for policy and language fit.
Raw prompt text, conversation history and the user's account record stay out of the partner ad-matching request.
Missing required permission or invalid configuration blocks delivery. Turning semantic targeting off does not bypass these checks.
01
Your app
Sends placement details and any permitted context to wavebird from your backend.
02
Firewall
Classifies permitted context, reduces the semantic signal and checks outbound fields.
03
Policy gate
Checks consent, categories, blocked industries, and route readiness.
04
Partner path
Receives only the scoped signal needed for a placement decision.
What teams can tell privacy reviewers
- Prompt sharing is off by default in the final API snippets.
- Semantic targeting requires explicit consent state in the request.
- Placement, transaction and regulatory fields are separate from semantic context.
- Without semantic consent, do not derive targeting from prompts. Permitted non-personalized product context remains subject to the active delivery rules.
A reduced signal is not a claim of anonymity
This describes the semantic signal, not every field in an ad request. A request also needs permitted placement, format, transaction and regulatory information. Additional field families are controlled by the integration's allowlist and policy; they are not automatically enabled.
Minimizing context reduces exposure. It does not make every technical identifier or derived field anonymous, and it does not replace the application's privacy notice.
Request logs, consent evidence and settlement records have separate purposes and retention rules. Review the privacy policy and the terms for your integration.
Continue from the boundary
Check the controls, then the implementation path.
Plan how ads fit your product.
Discuss placements, expected economics and data boundaries with your team. The integration guides and dashboard code give your developer the next step.