Our approach

TalkLayer is designed to collect only the account, configuration and conversation data needed to operate your website agent. This page is a product-ready draft and must be reviewed with counsel before commercial launch.

Account and workspace data

We expect to store profile details, organization membership, agent configuration, knowledge-source references, usage totals and billing status in the shared control plane.

Visitor conversations

Conversation transcripts, audio metadata and derived outcomes should follow the retention, consent and regional policies configured for each organization. Customers remain responsible for notices presented on their websites.

Security and subprocessors

Credentials remain server-side, account sessions use HttpOnly cookies, and browser embeds are bound to approved origins. A final policy will identify hosting, model, speech, analytics and billing subprocessors before launch.