Personalized AI for every member

Six changes shipped in this release. The assistant stopped being one-size-fits-all: each member’s role, goals, language, and context now travel into every interaction they have with it.

Personalization

• Two ways to set it up — on a first login with nothing saved, the assistant opens with role chips and runs a short interview, previewing what it learned before saving. A settings form covers the same ground for anyone who prefers to fill it in, and owners can edit a member’s setup from the team panel.
• Ten role presets — sales, account management, support, customer success, office management, marketing, owner, call center, analyst, and an open-ended option for a role that is not listed.
• It reaches the whole assistant — the member’s context flows into assistant chat and into both the summary and detailed briefings, sanitized per field and scoped to the organization.
• Gentle, not pushy — skipping onboarding keeps it closed for that session and offers it again next login until at least one field is filled.

Billing

• Rate per pack at a glance — the admin credit-pack cards show each pack’s effective price per credit, updating live in the form while it is being edited.
• Credits end to end — plan allowances are stored in credits internally, matching what customers already see. Identical usage produces identical deductions; the change decouples plan promises from conversion-rate changes.