Skip to main content

Voice hub orientation — the tile map for the voice desk

The voice docs split coverage by feature — one page per capability under Voice, plus a long guide list — because that is how the features shipped. Operators do not work that way. The question that brings someone to /voice is one sentence: where do I run the phone operations for this workspace? This page answers it, collecting the voice desk’s consoles into one walkthrough. Every console below is a tenant-owned control — you set it for your workspace, and nothing applies until you turn it on. Each tile maps 1:1 to a route under /voice in the dashboard.

Why the voice desk is one hub

A contact center is not a pile of pages — it is one queue that agents, supervisors, and admins each see from a different angle. The agent answers calls against a queue and marks dispositions; the supervisor watches the queue, jumps onto live calls, and coaches; the admin wires numbers, skills, and schedules that decide which agent the queue serves next. The voice hub holds all three angles together, so most tiles show the same call base sliced four ways: the live floor, the calling surface, routing plumbing, and evaluation + workforce. One question brings an operator here — where do I run the phones? — and the hub answers it in four rows.

The consoles

Supervisor floor and evaluation

These five consoles are where a supervisor lives during a shift. They share one queue and one call base, so they are best read together.

Contact center — /voice/contact-center

What it answers: what is happening on the floor right now — queue pulse, agent status, service level, live alerts, and callbacks? The supervisor’s home tile; the first stop at the start of any shift, and the surface the homepage triage links into. First visit: opens for owner and admin seats. Skip when you run a one-queue hobby desk — the wallboard covers it. Docs: Contact center live floor.

Live monitoring — /voice/monitoring

What it answers: which call do I join, listen, whisper, or barge right now? Real-time supervisor monitoring over active calls. First visit: open only when a supervisor seat operates the floor. Skip when you are the only agent on the team. Docs: Supervisor live monitoring (voice).

Wallboard — /voice/wallboard

What it answers: what does the whole floor see across every queue — calls waiting, average wait, service level, agents available, longest wait? The big-screen operations display. First visit: mounts onto the TV setup the moment the first queue exists; rewire the alarm thresholds before you hang it. Docs: Wallboard; Wallboard TV setup; Wallboard alarm rules.

Quality — /voice/quality

What it answers: which carrier leg is degrading — per-carrier MOS, jitter, packet-loss, and RTT, with drill-down to the worst calls of the last 7 days? Network and media health, not agent evaluation. First visit: read-only; wire an alert threshold once you know your normal. Skip when call quality is not on your incident list yet. Docs: Voice quality reading.

Evaluation hub — /voice/survey-scorecards

What it answers: where do scored evaluations against recorded calls live? The entry point for a QA program — the full lifecycle of scored evaluations sits behind this one tile. First visit: empty until a scorecard template exists. Docs: Survey scorecards; Voice QA scorecards; Quality evaluations.

Calling surface

Two surfaces let a browser act as a phone. Agents live in one of them; admins install the desk phones these complement.

Browser softphone — /voice/softphone

What it answers: how do I place and receive calls from the dashboard itself? Registers your browser as a SIP endpoint over a secure WebSocket, no desk phone required. First visit: the agent’s home tile — open it first, stay here. Skip only when your agents run hard phones. Docs: Browser softphone; Softphone browser calling.

Desk phones — /voice/devices

What it answers: how do I provision a SIP desk phone onto my extension — the server-level setup, the provisioning profile, the assigned extension? The admin surface that hands a physical phone the tenant’s SIP edge; provisioning is minted entirely server-side. First visit: optional — skip when the team is softphone-only. Docs: Desk phone provisioning; Extensions.

Routing and queue plumbing

Numbers, skills, and text-back decide which agent the queue serves next. Admins wire these; agents and supervisors consume the outcome.

DNIS routes — /voice/dnis-routes

What it answers: how do N inbound numbers funnel into one queue via a prefix or regex pattern, instead of one inbound rule per DID? The many-to-one inbound management surface for the matching engine. First visit: skip when you run a single published number — the inbound route editor covers it. Docs: DNIS pattern routing.

Call park lobby — /voice/park

What it answers: where did a parked call go, and who retrieves it? The park and retrieve surface agents use mid-call, because hold-by-slot beats blind transfer when a warm handoff is not ready. First visit: skip when your team is one agent — you will never park a call. Docs: Call park lobby; Paging, ring groups, and call park.

Extension SMS — /voice/extension-sms

What it answers: how do I text back from the business number as the extension that just handled the call? Per-extension two-way texting — the thread an agent continues after a voice contact. First visit: skip when your tenant has no SMS-enabled voice DIDs — the tile is inert without one. Docs: Extension SMS; Extension identity model.

Scheduling and CSAT/NPS analytics

Workforce scheduling — /voice/scheduling

What it answers: who works when — shift rosters, the staffing forecast, and the swap / time-off requests agents submit? The WFM surface the roster runs on. First visit: skip while the team is under five agents — a shared calendar beats a WFM rollout. Docs: Voice calendars; Callbacks scheduling.

Post-call CSAT — /voice/post-call-csat

What it answers: how satisfied was the caller, on a 1–5 survey played right after the call? The CSAT IVR survey and its analytics. First visit: pick either CSAT or NPS; most teams run one survey, not both. Docs: Auto CSAT/NPS surveys; CSAT/NPS in Live Ops settings.

Post-call NPS — /voice/post-call-nps

What it answers: how loyal is the caller base, on a 0–10 NPS survey played right after the call? The NPS IVR survey and its analytics. First visit: as with CSAT — run one survey unless you genuinely report both measures. Docs: Auto CSAT/NPS surveys; Auto NPS settings.

Skill governance

Skill catalog — /voice/skill-catalog

What it answers: which routable skills exist in this workspace — language, product, region, escalation tier? The skill taxonomy you queue-route against. First visit: admin-only; define the catalog before certifications point at it, because certifications hang off catalog entries. Docs: Skill catalog; Skill catalog guide.

Skill certifications — /voice/skill-certifications

What it answers: which agent holds which skill, at which proficiency, with whose sign-off? The assignment map routing consumes. First visit: skip when every agent takes every call — an empty catalog never routes. Docs: Skill certifications.

Role gating

The /voice sidebar group renders for every seat, but the tiles gate differently. Row-level role checks live on the page; the backend re-checks send scopes, so the UI gate is an affordance, never the boundary.
  • Agent seat — opens Softphone, Park, Extension SMS, and Scheduling (self view). The calling surface is the agent’s home.
  • Supervisor seat — adds Monitoring, Wallboard, Contact center, and the evaluation consoles. Read-mostly on routing; a supervisor can escalate into a live call from Monitoring.
  • Admin / owner — plus DNIS routes, Devices, Skill catalog, Skill certifications, and the CSAT/NPS survey configs. The plumbing tiles respond to admin changes only — nothing applies until an admin wires it.
Tenant isolation applies across every tile: queues, DIDs, skills, and certifications live in your workspace only, and nothing crosses tenant boundaries.

Suggested rollout order

Take the tiles in this order when you stand the desk up for the first time:
  1. Softphone or Devices — give agents a way to answer. Skip Devices when softphone-only.
  2. DNIS routes — funnel the numbers into one queue. Skip when a single DID serves the desk.
  3. Skill catalog, then certifications — only if routing by skill matters. Skip while every agent takes every call.
  4. Wallboard, Contact center, Monitoring — the supervisor loop. Skip while the team is one agent.
  5. Quality — wire the network-health threshold once traffic is real.
  6. Survey scorecards — start the QA program; skip until a scorecard template exists.
  7. Post-call CSAT or NPS — pick one survey.
  8. Scheduling — last, once the roster is too big for a calendar.
Steps 1–4 put the floor on the air. Steps 5–8 layer management on top; add them as the hallway grows, not before.

Hub tiles at a glance

Contact center

Supervisor floor: queue pulse, agent status, service level, and live callbacks at a glance.

Live monitoring

Join, listen, whisper, or barge onto an active call as it happens.

Wallboard

Big-screen operations display across every queue — waiting, service level, longest wait.

Browser softphone

Register the browser as a SIP endpoint and place or receive calls in the dashboard.

DNIS routes

Funnel N inbound numbers into one queue via a prefix or regex pattern.

Skill catalog

The routable-skill taxonomy queue routing consumes.

Workforce scheduling

Shift rosters, staffing forecast, and the swap / time-off requests agents submit.

Quality

Per-carrier MOS, jitter, and packet-loss with drill-down to the worst calls.

Where the in-depth docs live

  • Voice group — the per-capability reference pages under the Voice nav are the detailed provider contract for each console.
  • Guides — the granular walkthroughs linked per tile above; the Guides nav groups them by task rather than by hub tile.

See also