Quick answer: Workforce management (WFM) on Devotel Orbit is the closed loop that turns a volume forecast into staffed shifts and then measures — in real time, against the plan — whether agents are where the plan said they would be. It is built into the contact-center pillar, not bolted on as a separately licensed module: the same tenant schema that stores queues and conversations also stores forecasts, shift assignments, adherence events, and exception approvals. This post walks the entire shipped WFM surface end to end, names the operating patterns it supports, and is honest about where a dedicated WFM specialist still goes deeper.
1. The closed loop: forecast to schedule to adherence
WFM answers three questions in sequence, and a tool that stops early leaves the loop open:
- How much work is coming? — the forecast, per queue and channel, compared against actuals so accuracy is measured rather than assumed.
- Who covers it? — shift definitions, per-agent assignments, and the shrinkage and occupancy assumptions between headcount and workable hours.
- Did the plan hold? — real-time adherence against scheduled activities, with exception workflows that keep the report honest when reality intervenes (outages, emergency meetings, approved absences).
Each stage feeds the next. A forecast nobody schedules against is a chart; a schedule nobody measures adherence against is a wish; adherence with no exception path punishes agents for a dialer outage they did not cause. The sections below walk the shipped surface in that loop order.
2. The shipped WFM surface, module by module
Every capability below is live in the dashboard today, scoped per tenant. Each subsection names the workflow it carries rather than an internal module — the map a buyer needs is workflow-to-screen, not workflow-to-filename.
2.1 Forecasts, shift templates, and assignments
The scheduling module is the planning core. It holds volume forecasts per channel, a forecast-accuracy readout (planned versus actual, so the forecaster learns), shrinkage assumptions (the share of paid time lost to absence, breaks, and unplanned work), and shift templates a scheduler applies to agents as assignments — individually or in bulk, updatable and deletable as the week changes. Assignments carry a channel dimension, so a blended voice/chat team can be scheduled against the channel each shift actually serves.
2.2 The daily summary
A summary view rolls the plan for a given date into one screen — how many agents are scheduled, which queues they cover, and how the day is trending against the forecast. It is the morning-standup artifact: open it, compare coverage to expected volume, and know before the first queue interval whether the day starts staffed or short.
2.3 Real-time adherence
The adherence module compares each agent's current activity against their scheduled activity and flags divergence as it happens, not in a weekly report. It ships with a default adherence target of 85% — a deliberate choice, since a 100% target treats every legitimate break as a failure — and a configurable breach threshold (1–480 minutes, default 5) so a supervisor can separate "agent is 90 seconds late back from break" from "agent has been off-phone for forty minutes." Breaches surface as a live list a team lead works down during the day.
2.4 Team adherence rollup
For the manager who does not watch individual agents, a team-level view computes adherence percentage per queue per day — the aggregate number that answers "is this queue following its schedule" without naming names. It takes an optional queue filter, so a multi-queue operation reads each line of business separately.
2.5 Adherence exceptions with approval
Real operations have legitimate divergence: an outage pulled everyone into an emergency meeting, a coaching session ran long, an absence was approved after the schedule published. The exceptions module lets a supervisor file an exception — typed as system_outage, emergency_meeting, unplanned_coaching, approved_absence, training, or other — and routes it through an explicit approve / deny / cancel lifecycle. Approved exceptions stop counting against adherence, which is what keeps the adherence report honest: it measures divergence from plan minus the divergence the business decided was legitimate. A trends view counts the exception population over time (the queue itself is capped at 200 rows per page, with an uncapped total so "showing N of M" always reconciles).
2.6 Shift activities and conformance
A shift is not a monolith — it is lunch, two breaks, a coaching block, and the queue work around them. The shift-activities module breaks each shift into typed activities (lunch, break, training, coaching, meeting, off_phone, other) with local start and end times, so adherence compares agents against the activity they should be doing right now, not a generic "on shift" state. A conformance view shows how closely the day's actual activity timeline matched that plan.
2.7 Intraday staffing
The morning forecast never survives contact with the day. The intraday staffing view recomputes required versus actual staffing interval by interval for the current date — the reforecasting surface. When an interval trends short, the answer comes from the next three modules.
2.8 Open-shift bidding
When coverage is short, a scheduler can publish an open shift to the team instead of assigning it. Agents who want the hours submit bids, can withdraw them, and the scheduler awards the shift to a chosen bid. Published shifts can be cancelled if the need clears. The workflow keeps extra-hours distribution transparent rather than a manager's private negotiation.
2.9 Overtime offers
Overtime is the same market mechanism pointed the other way: the scheduler publishes an overtime offer, eligible agents submit claims (withdrawable until awarded), and the scheduler awards the offer. Claims give the scheduler a pool to choose from instead of a cold call down a phone list — and every step is recorded, so overtime distribution is auditable.
2.10 VTO (voluntary time off) offers
When volume falls below forecast, the symmetric workflow: publish a VTO offer, agents claim it, the scheduler awards, and the released hours come off the plan. Intraday surplus management without the awkward "does anyone want to go home early" broadcast.
2.11 Timekeeping and payroll export
The timekeeping module closes the loop at the end of the pay period: per-agent timesheets and a payroll export (CSV download) so approved hours, overtime awards, and exception-backed absences reach payroll as data, not as a spreadsheet someone retyped.
2.12 Agent skills
Skill records — an agent's proficiency per skill, upserted and removable — feed the contact center's skill-based routing. WFM uses the same source: a schedule built on outdated skill data routes work to people who cannot take it, so skill maintenance lives inside the WFM surface where schedulers actually work.
3. Operating patterns the surface supports
Three patterns recur once the loop above is wired together:
- Intraday reforecasting. The day's plan is a hypothesis. The intraday staffing view recomputes interval staffing as actuals arrive; the response levers — overtime offers for shortfall, VTO offers for surplus, open shifts for tomorrow's gap — are first-class workflows, not manual outreach.
- Skill-based coverage. Routing and scheduling read one skills surface. A scheduler assigns shifts against the channel and queue an agent is actually skilled for, and the routing side honors the same records.
- Non-punitive adherence. The 85% default target, the configurable breach threshold, and the typed exception lifecycle exist so adherence measures plan-versus-reality without taxing agents for legitimate divergence. Teams that skip the exception workflow end up with agents avoiding the phone to protect their numbers — the workflow is the feature.
4. Honest parity: built-in WFM versus the specialists
Dedicated WFM vendors — the workforce-management suites sold alongside or inside enterprise contact-center deployments — cover the same loop with more exotic mechanics: long-horizon simulation scenario modeling, deep historical pattern libraries, and agent-facing mobile apps decades in the making. If your contact center runs thousands of agents across many sites with a full-time WFM analyst team, that specialist depth is real and worth paying for.
For the operations buyer Orbit serves, the parity question is different: the built-in WFM surface covers the entire daily loop — forecast, accuracy, shrinkage, schedule, intraday, adherence, exceptions, market-style shift and OT/VTO distribution, timekeeping, skills — without a second contract, a second vendor integration, or a second system of record to reconcile. The two specialist capabilities Orbit does not claim to replicate are agent-owned self-scheduling negotiation (shift-swap markets with rule engines) and multi-site simulation scenario modeling; both are listed here rather than silently absent. Evaluate the loop first and the exotic features second — most stalled WFM programs fail on the loop.
5. Where to look next
The pillar page that owns this subsystem is Contact center (CCaaS), which lists workforce management among its shipped capabilities; the Customer experience (CXaaS) pillar carries the same WFM surface into the feedback-and-quality narrative. For category-scale vendor context, the contact center comparison ranks the seat-licensed incumbents, and the sibling WEM/WFM buyer's primer maps the wider workforce-engagement umbrella this WFM loop sits inside.
Because scheduling, adherence rules, exception approvals, and payroll exports are per-tenant operations, every control described above is configured and audited within your own tenant — Orbit hosts the surface; your operations team owns the posture.