Admin console

#Websites & widget

Manage brands and websites here: add a brand (name, slug), register a website (brand, domain, extra origins), copy the install snippet and reveal the player identity signing secret. The Widget appearance page sets colour, position, theme, default language, texts per language, the satisfaction survey and the outside-hours notice. Details: Chat widget, Player identity.

On Proactive triggers you create, enable, disable and delete triggers that show visitors a message automatically under given conditions.

#Team & access

Three tabs:

• Members: Invite member (e-mail, name, roles, departments). An invite produces a one-time link that you pass on yourself. You can edit a member’s roles, departments and concurrent chat limit, create a new invite or password-reset link, and remove the member from the tenant (their open conversations go back to the queue).
• Roles: system and custom roles; see Roles.
• Departments & hours: the tenant time zone and weekly business hours (or 24/7). Departments have a name, description, their own schedule, members and websites; new conversations from a website start in that website’s department.

#Analytics

Tabs: Overview, Live, Conversations, Agents, CSAT, Actions and Visitors. Pick a period (today, 7/30/90 days or custom), website and department; data can be exported as CSV.

• Key figures: AI resolution rate, escalation rate, abandonment rate, median first response, queue wait and resolution time.
• CSAT: average score, score distribution and comments.
• Live: queue, longest wait and pending offers.

#Integrations, actions and approvals

Integrations holds the operator API connection; Bonus campaigns the campaigns the AI and agents may offer (operator campaign id, amount, wagering multiplier, TR/EN terms, new-player and minimum-deposit rules, validity dates); Deposit methods fallback deposit methods; and Action policy bonus limits and the approval threshold. Details: Operator API.

Actions & approvals has two tabs: Log (filter by status, player id and conversation id) and Approval queue. With actions:approve you approve or reject waiting requests with a note, which is kept in the audit log.

#Telegram, API keys and webhooks

• Telegram & routing: waiting queue, agent status, routing settings, bot and staff groups. See Telegram bridge.
• API keys: server-to-server access. See REST API.
• Webhooks: endpoints, event selection, test sends and the delivery log. See Webhooks.

#Operations & incidents

Viewing needs operations:read, changes need operations:manage. You can be notified of incidents opening and closing through a webhook.

#Usage & billing

Pick a month to see its usage: conversations, AI-resolved conversations, agent conversations, messages, AI tokens, active agents, Telegram messages, operator actions and webhook deliveries. When prices are set you also see an estimated charge, the last 6 months and a 12-month CSV export (billing:read).

#Privacy & telemetry

• Telemetry policy: retention (1–730 days), sample rate, stored IP address (truncated, full or none), require consent and honour Global Privacy Control. The consent snippet is on this page too.
• Conversation retention: anonymise closed conversations after 7–3650 days, or keep them indefinitely. Anonymising removes texts, notes, files and comments; counts, times, statuses and ratings are kept for analytics.
• Erase a visitor: anonymises all of a visitor’s conversations and deletes their files (you retype the id to confirm; visitors:erase). There is a separate action to erase only the visitor’s analytics.

#Security & audit log

Filter the audit log by action, actor type (user, AI, system, API key), actor, target and date range, and open any entry for details. Verify integrity recomputes the chained hashes of the entries to check that the log has not been altered (audit:read).