Skip to main content
Conversational context is sensitive. Adgentek processes the context supplied in an ad request to derive signals used for matching and decisioning. Publishers should send only the context needed for that request and must not send raw personally identifiable information.

​
What is transmitted to demand partners

  • content categories and keywords
  • intent, sentiment, funnel-stage, language, and engagement signals
  • extracted entities and brand signals
  • standard advertising request data when available, such as device, user-agent, IP, geography, and approved identifiers
Raw conversation messages are not included as a dedicated field in the current demand-partner bid request. That distinction does not make all bid data anonymous: standard request fields can still be personal data under applicable law.

​
Processing and storage

Do not describe raw context as immediately discarded or processed only in memory. The current semantic-processing path can store supplied text for embedding and cache operations. Retention, lawful basis, consent, regional requirements, and data-subject rights are governed by the applicable agreement and privacy notices—not by this technical summary. Before launch, review the exact fields your integration sends, avoid unnecessary identifiers, and apply your consent and disclosure requirements.

​
Data processing terms

Data processing terms, subprocessors, and regional details are covered in the Privacy policy and the Subprocessors list. For DPA execution, contact hello@adgentek.ai.