Convoship Docs

Product documentation

Workspace settings

Plan, AI agent production defaults, the workspace model, log privacy, MFA, and session security.

Workspace → Settings holds workspace metadata, the workspace-wide AI agent controls, and security settings for your account in this workspace.

General

  • Workspace name — display name for the team container.
  • Plan — current subscription tier and session limits (see usage meter in the sidebar).
  • Your role — owner, admin, developer, editor, or viewer in this workspace.

AI Agent model

Choose the model every AI Agent in this workspace runs on. Default is Convoship's managed Claude. Admins can switch to their own Anthropic key (BYOK) or a custom OpenAI-compatible endpoint — for example a fine-tuned small language model, Convoship-hosted or in your own infrastructure. Includes a Test connection check and an optional, off-by-default fallback that retries failed custom-endpoint turns on Claude. Fine-tuned models are created in the Models section (see Workspace models docs).

AI agent production defaults & log privacy

The same page carries the workspace-wide agent guardrails — per-agent rate limits, the eval publish gate, publish approval, cost-per-contact (see Production settings docs) — and the PHI/PII conversation-log redaction toggle (see Conversation log privacy docs).

Multi-factor authentication

Enable TOTP-based MFA on your user account. Once enabled, login requires your authenticator app in addition to password.

Sign out everywhere

Revokes all active sessions for your user across devices. Use this if you suspect credential compromise.