!AI flagged Thu 14:00–16:00 · Chat (−2 agents, vol +12% over forecast) and Tue 12:00 · Voice (−1). Click any cell for interval detail and an Auto-Pilot recovery move.
Forecast vs Actual vs Scheduled
ForecastActualScheduled±2% band
Staffing Requirement Curve
Erlang C
Required HCScheduled HCGap
SLA by Queue
SLA Breach Prediction
Queue
SLA %
Target
Breaches
ASA
Offered
Trend
Predicted
AI
Breach prediction: Chat projected to dip to 92% in the 13:00–16:00 block (78% confidence). Pre-empt by adding 2 agents or shifting Email volume.
Coverage Gap · Now
Watch
−2
Chat · 14:00–16:00
Required 14 · scheduled 12 · actual 11. Other intervals covered.
Real-Time Adherence
92% team
Agent
Sched
Actual
Var
Adh%
Capacity Levers
OT/VTO Optimizer
Shrinkage vs Plan
28%
▼ from 31% · target ≤30% · teal
Overtime Budget Burn
38 hrs usedbudget 60 hrs · 63%
Occupancy
81%
Utilization
74%
AI
Cheapest close for 13:00–16:00 gap: offer VTO on Email (over by 3) + move 2 to Chat. $0 OT, projected SLA 96%.
Forecast Self-Tuning
Nightly reconcile
Yesterday accuracy94.1%
Worst interval (MAPE)13:00 · 9.2%
Best interval09:00 · 1.1%
Drift vs 7-day mean▲0.4pp
Model retrained02:14 ✓
Field GuideHow a WFM analyst runs the floor with Cadence
How to use this view
1Start at the heatmap, not the KPIs. A green-dominant grid means the week is staffed; scan for amber/red islands first — those are the only intervals that need a decision today.
2Read a cell as a ratio, not a number. Color = actual+scheduled coverage ÷ Erlang requirement. Teal ≥100%, amber 90–94%, red <90%. The "!" marks AI-confirmed breach risk.
3Confirm the cause before you move people. Click an amber cell — the Interval modal splits root cause into volume spike vs understaffing. The fix differs: surge wants a temporary move, structural understaffing wants a schedule change.
4Let the optimizer cost the fix. The OT/VTO Optimizer ranks close-the-gap options by dollars — a skill move or VTO on an over-covered queue is almost always cheaper than overtime.
5Trust but reconcile. Forecast Self-Tuning reports accuracy drift by interval nightly. If one interval keeps missing, the forecast is wrong — not your agents.
6For your own org: do you plan to a single daily SLA, or to the worst 30-minute interval? Intraday coverage is won or lost in the red cells, not the average.
Watch the Walkthrough
Four AI agents walk this dashboard.
In Context · Staffing Signals
Sample Feed
Illustrative — wire to your WFM / HRIS & ACD feed.
Next-hour forecastinbound contacts, all queues418▲ 6%
Logged-in vs requiredcurrent interval headcount37 / 39▼ 2