The fastest fixes, grouped by where you noticed the problem.
Access and sign-in
| Symptom | Fix |
|---|
| Cannot sign in | Ask an admin to resend your invite from Admin -> Invites |
| No Admin entry in the sidebar | You do not have the admin role, see People and roles |
| Location dropdown is empty | You have no location grants, see Location permissions |
| A page 403s | The API re-checks permissions; the grant is genuinely missing, not just hidden |
Forecasting
| Symptom | Fix |
|---|
| Generate is disabled | Fill in location, week, and projection model |
| Forecast is $0 everywhere | No historical actuals; import history or use Trailing Average, see Actuals and variance |
| A day has no slots | Operating hours are missing for that weekday, see Location hours |
| Cannot edit slots | The forecast is Pending Approval or Published and is locked by design |
| Labor generate button missing | Select a Published sales forecast |
| Staffing is zero at peak | The top labor matrix band has an upper bound; leave "volume to" empty |
| Labor cost is zero | The job position has no hourly rate |
| Nobody can approve | Add approvers in Approval settings |
Dashboards and data
| Symptom | Fix |
|---|
| Dashboard is empty | Data store inactive, not default, or the location is not mapped to a store key, see Data stores |
| A new warehouse table is missing | Click Clear cache on the data store row |
| Custom dashboard widget goes blank after filtering | The widget's data lacks the field named in filters |
| Embed shows nothing | Regenerate the token; it may have expired, see Embedding dashboards |
Notifications
| Symptom | Fix |
|---|
| A rule never fires | Org override may be Off for everyone, or nobody has it enabled |
| One user receives nothing | Their subscription is off, muted, or inside quiet hours |
| Slack silent | Invite the bot with /invite @UnitOps and re-run the channel test |
| Email silent | Verify the provider and From address in Notification channels |
| Push silent | Re-enable push on the device in Settings -> Notifications |
| Deliveries stopped abruptly | Check Admin -> Notifications -> Delivery Health for the failure reason, then replay after fixing the cause |
| Too many alerts | Raise thresholds on the role default, lengthen cooldown, or switch email to a digest |
| Symptom | Fix |
|---|
| "AI is not configured" | The platform AI configuration is unavailable; contact whoever operates your deployment, see AI configuration |
| "Test connection" fails in Admin -> AI Config | The platform AI configuration is unavailable or misconfigured; contact your operator |
| Chat or summaries stop with a budget error | The monthly budget was reached; raise it in Admin -> AI Config, see AI configuration |
| Documents are ignored | Index the document; document search also needs the platform embedding provider, see Document storage and embeddings |
| Answers are generic | Name the location and the period in your question |
Still stuck
Collect the page URL, the exact error text, the time, and the affected location, then email support@datalakehouse.io. If the problem is configuration-related, reproduce it on staging first: https://dev-staging-unitops.datalakehouse.io.