Skip to content

📊 Metrics > Realtime

The Realtime page gives supervisors a live operational view of Kewbot: current conversation load, team distribution, and agent performance.

Note: some labels or columns may vary depending on tenant configuration, modules enabled, and user permissions.

Placeholder - Dashboard Tiempo real

What this page is for

  • Detect bottlenecks early (growing queue, too many unassigned conversations).
  • Identify overloaded or idle agents.
  • Make quick decisions: reassign, prioritize, escalate.
  • Monitor operations during shift changes or incident peaks.

Filters and controls (top section)

Field / control What it shows Why it matters
Range / period Time window for current view (e.g., last X minutes). Focus on immediate operations or short-term trends.
Group Teams configured in Kewbot. Analyze one team’s workload and performance.
Agent A specific operator user. Audit individual activity and load.
Channel Source channel (according to your setup). Compare behavior by channel.
Refresh / Auto-refresh Automatic update status and manual refresh option. Keep metrics current without full page reload.
Last updated Timestamp of latest dashboard refresh. Validate that data is up to date.

Main KPI cards

KPI What it represents How to interpret it
Active conversations Total conversations currently in progress. A rapid increase may signal upcoming saturation.
Unassigned Conversations not yet taken by an agent/team. If this grows, review routing rules and staffing.
Pending / waiting Open conversations still unresolved. Useful to detect operational accumulation.
SLA breached / out of target Cases exceeding target response/handling time. Prioritize these to reduce service risk.
Agents online Agents currently connected and available. Compare available capacity vs incoming demand.
Occupancy Relationship between active load and team capacity. Spot overload or underutilization quickly.

Agents table (operational detail)

Column What it shows Practical use
Agent Operator name. Identify who is handling what.
Status Available, busy, away, etc. Decide where to route new workload.
Group Assigned team. Validate team-level balancing.
Active conversations Number of current cases handled by the agent. Detect individual overload.
New assigned Recently assigned/taken conversations. Track recent intake per agent.
Avg response time Average response latency for the agent. Monitor service speed.
Last activity Timestamp of latest interaction. Detect inactivity quickly.

Conversations table (live traceability)

Column What it shows Practical use
Conversation ID Unique case identifier. Auditing and traceability.
Contact / customer Contact name or identifier. Locate conversations fast.
Status Open, in progress, pending, closed, etc. Understand service stage.
Assigned to Current responsible agent. Confirm conversation ownership.
Group Team handling the conversation. Verify operational distribution.
Last message Time/type of latest conversation event. Prioritize oldest silent conversations.
Queue time / age Time since intake or last key status change. Aging-based prioritization.
Tags Labels applied to the conversation. Segment by topic, priority, campaign, etc.
Source Campaign, inbound route, ITR, or entry source. Diagnose performance by source.

Common actions from Realtime

  • Open conversation for full context.
  • Reassign across agents or groups.
  • Take conversation (if enabled by flow and permissions).
  • Apply fast filters during incidents (team, status, aging).

Supervision best practices

  1. Review Unassigned + SLA breached first.
  2. Balance workload by agent before mass reassignment.
  3. Combine Realtime view with tags and conversation source.
  4. At shift close, ensure aged conversations still have clear ownership.