User guide

SmartHub guide

How to use SmartHub: inbox, AI drafts, reviewing and approving replies, attachments, queues, reports, users, settings, and troubleshooting.

← Back to Help & FAQ

This manual covers the SmartHub (Operations Hub) interface for support agents, supervisors, and administrators who review inbound conversations, generate and approve safe AI replies, manage queues, and export reports.

SmartHub is SmartQueue's optional human-in-the-loop workspace. SmartQueue can answer email automatically; SmartHub is where your team reviews, approves, and sends replies when you want a person in the loop, and where you handle anything that was escalated.

Included:

  • Sign-in, roles, and the Inbox.
  • Reading cases, updating fields, AI drafts, and sending replies.
  • Attachments, internal notes, queues, reports, users, and settings.
  • Privacy rules, troubleshooting, and a safe-operating checklist.

Not included:

  • API/technical documentation and deployment tasks.

1) Access and sign-in

Open the SmartHub URL provided by SmartQueue and sign in with your assigned email and password.

Access rules:

  • If your organization or user account is inactive, SmartHub will not allow access.
  • If your account is active but the AI budget is disabled or exhausted, you can still review cases, but AI draft generation is unavailable until an administrator updates the budget.

Roles (what each can do):

  • Agent: Inbox, case fields, replies, attachments, and internal notes.
  • Supervisor: Agent permissions plus Queues and Reports.
  • Admin: Full organization access, including Users and Settings.

Tip:

  • Always Log out on shared or unmanaged devices.

2) Inbox overview

The Inbox is the main work area. It shows:

  • Queue selector: all queues or a specific queue.
  • Status counters: All, Open, Pending, Closed, and Resolved.
  • Inbox search: find cases by case ID, contact, email, subject, company, summary, tags, and message text.
  • Time-frame selector: all, today, yesterday, last 7/30 days, this month, or a custom range.
  • Conversation list: the cases in scope.
  • Case detail: the selected conversation, fields, messages, attachments, AI draft, notes, and resolution controls.
  • Right context panel: case profile, AI signals, and timeline.

3) Filtering the inbox

Filter before taking action so you work from the correct case set.

  1. Select a queue, or keep All queues.
  2. Select a status counter (use All for no status filter).
  3. Type in Search inbox to narrow by details or message content.
  4. Select a time frame; for Custom range, choose From and To dates.
  5. Use Clear to reset filters.

Note:

  • If no cases match, the list shows an empty state rather than hiding the controls.

4) Reading a case

Open a case from the conversation list and review:

  • Subject and case ID.
  • Contact, email, company, source, queue, owner, SLA, priority, tag, and sentiment.
  • The full message history.
  • Attachments and download status.
  • AI signals and the timeline.

Important:

  • Always read the latest customer message before approving a reply.
  • Do not rely only on the summary or tags when a case is sensitive, legal, billing, cancellation, refund, or contains personal data.

5) Updating case fields

Agents can update operational fields from the case detail panel:

  • Status: Open, Pending, Closed, Resolved.
  • Queue, Tag, Priority, Assignee.

Changing a field saves the updated Hub state. If your role does not allow an action, SmartHub blocks the save.

Tip:

  • Use Pending for cases that need human follow-up so they are easy to find later.

6) AI drafts

Use Generate answer with AI to request a draft for an email case.

Depending on your organization settings, the Hub may draft from:

  • The SmartQueue runtime.
  • The Hub reply library.
  • A hybrid of both.

AI budget rules:

  • If the organization budget is disabled or exhausted, draft generation is blocked.
  • Budget is managed by SmartQueue administrators.
  • You can still review cases manually when AI generation is blocked.

Tip:

  • Use Regenerate only after you change context or review a poor suggestion — it consumes additional budget.

7) Reviewing and sending replies

Before resolving or approving a reply:

  1. Read the latest customer message.
  2. Check attachments if the customer references a file.
  3. Review the draft for accuracy, tone, and policy safety.
  4. Edit or replace the draft if needed.
  5. Confirm the correct status and owner.
  6. Resolve or save the case only when the response is ready.

Note:

  • In staging/test environments, delivery may be rerouted to approved test recipients.
  • In live production, approved replies may send customer-facing email depending on the configured mode.

8) Attachments

Attachments appear under the message thread.

  • Download is available when the Hub has the attachment bytes.
  • Content unavailable means the metadata exists but the original bytes were not stored.
  • If you have the original file, use the restore/upload control to attach the content before review.

Security:

  • Treat all customer files as sensitive. Do not download or forward attachments unless your work requires it.

9) Internal notes

Use internal notes for context that should not be sent to the customer:

  • Why a case was escalated.
  • What was checked and what a human should verify.
  • Delivery or attachment issues.

Do not include unnecessary personal data, secrets, passwords, or payment details in notes.

10) Queues

Supervisors and admins use Queues to manage routing:

  • Queue name and purpose.
  • Routing mode and SLA/fallback behavior.
  • Members assigned to the queue.
  • Queue-specific open workload.

Use View queue to jump to the Inbox filtered to that queue.

11) Reports

Reports help supervisors and admins audit work and export case data.

Filters include search, period/custom range, person/owner, queue, status, tag, channel, priority, source, sentiment, SLA, company, and case flag.

Use Export Excel/CSV to save a report for the current filtered set. The export requires sign-in and Reports access.

12) Users

Admins can manage Hub users:

  • Name, Email, Role, Status, and Queue assignments.
  • Set a user to Inactive when they should no longer have access. Inactive users cannot sign in, and existing sessions stop working once the backend sees the change.

13) Settings

Settings contains operational configuration and readiness information:

  • Forwarding status.
  • Hub integration/API key status.
  • AI draft mode.
  • Reply library and approved-reply behavior.
  • Production readiness checks.

Security:

  • Do not share API keys. If a key may have been exposed, ask a SmartQueue administrator to rotate it.

14) Privacy and security rules

Follow these for every case:

  • Verify the recipient and case before sending or resolving.
  • Never paste passwords, payment card data, government IDs, or secrets into notes or drafts.
  • Do not download attachments unless necessary.
  • Do not forward customer data outside approved channels.
  • Use Pending when a case needs human review.
  • Escalate unclear, legal, billing, cancellation, refund, medical, security, or personal-data cases.
  • Log out when finished.

15) Troubleshooting

Cannot sign in:

  • Confirm email and password.
  • Confirm the organization and user are active.
  • Ask an admin to check billing/access status.

AI draft button is blocked:

  • The account may have no AI budget or may have exhausted it.
  • The case may not be an email case.
  • The AI runtime may be in a hold/review state.

Search does not show expected cases:

  • Clear filters, switch status to All, set time frame to All time, and confirm the correct queue.

Attachment will not download:

  • It may be metadata-only — restore content if you have the original file.
  • Ask an admin to inspect the inbound import.

Report export is empty:

  • Clear report filters, expand the date range, and confirm your role includes Reports access.

16) Safe operating checklist

Before using SmartHub with real customers, confirm:

  • The organization is active and billing/access is correct.
  • The AI budget is set.
  • Forwarding is verified.
  • Test/live mode expectations are clear.
  • Allowed users and roles are correct.
  • Queues and SLAs are configured.
  • Report export access is set.
  • The escalation policy is understood by the team.

Need a hand?

Our team can walk you through setup in a quick call.

Book a demo