Best practices

Practical tips by area — dashboard, chat, templates, meetings, knowledge, and projects.

Dashboard (AI Staff)

  • Open AI Staff at the start of the week to spot usage trends and work that may need follow-up.
  • Use department views to see which teams are active — share template wins with admins when something works well.
  • Treat dashboard numbers as signals, not scores; open Chat or Meetings minutes when a metric looks unusual.

Chat

  • State your goal, time range, and output format in the first message (bullets, table, email draft, etc.).
  • Refine answers with follow-up prompts instead of starting a new chat from scratch.
  • Check source links in every response before sharing outputs with clients or leadership.

Task templates (New task)

  • Pick a template that matches the job — summary, extraction, or draft — rather than writing a blank prompt.
  • Fill every field with specific names, dates, and file references; vague inputs produce vague outputs.
  • Reuse the same template for recurring work (weekly reports, invoice checks) so outputs stay consistent.

Meeting records (Meetings minutes)

  • Give meetings clear, searchable titles with date and topic before upload or recording.
  • Wait for processing to finish, then review action items and owners before asking Chat about the same meeting.
  • Push important meetings to Knowledge when you want them cited in future answers across the workspace.

Knowledge

  • Upload focused files by topic and use descriptive names so AI can find the right source quickly.
  • Confirm upload status shows complete before asking Chat questions that depend on new documents.
  • Remove outdated versions and keep sensitive files in approved folders so answers do not cite stale policy.

Projects

  • Group related chats into a Project so context stays together across sessions.
  • Name projects after the client, initiative, or quarter — not “Untitled” or “Test”.
  • Archive finished projects so your sidebar stays easy to scan.