Skip to main content

Dashboard

The VOCALS dashboard provides a real-time analytics overview of your call operations. It surfaces the metrics that matter for monitoring call quality, cost, and agent performance.

Overview​

When you log in, the dashboard displays a summary of your key metrics for the selected time period. The dashboard is organized into tabs:

  • Overview — Key metrics at a glance (Total Calls, Total Minutes, Avg Duration, Completed %, Failed %, Avg Cost/Call, Barge-in Rate)
  • Performance — Latency and pipeline performance metrics
  • Costs — Cost breakdown by provider and over time
  • Agents — Per-agent performance comparison
  • Custom — Customizable dashboard view

Each metric is shown as a card with the current value. Use the All Agents dropdown to filter by a specific agent.

Available Metrics​

Call Volume​

  • Total calls: Count of all calls (inbound + outbound) in the selected period.
  • Inbound vs. outbound breakdown: Split view showing call direction distribution.
  • Calls over time: Line chart showing call volume by hour, day, or week depending on the selected range.

Latency​

Pipeline latency is broken down by stage so you can identify bottlenecks:

  • STT latency: Time from audio received to transcript available (per utterance).
  • LLM latency: Time from transcript sent to response generated (time to first token and total).
  • TTS latency: Time from text sent to audio playback started.
  • Total round-trip latency: End-to-end time from when the caller finishes speaking to when they hear the agent's response.

Latency is displayed as p50, p90, and p99 percentiles. The goal for a natural-sounding conversation is under 800ms total round-trip at p50.

Cost​

  • Total cost: Aggregated cost across all providers and Twilio usage.
  • Cost per call: Average cost broken down by STT, LLM, TTS, and telephony.
  • Cost by provider: See which providers are driving your spend.
  • Cost trend: Daily cost chart to spot unexpected spikes.

Conversation Metrics​

  • Conversation turns: Average number of back-and-forth exchanges per call.
  • Barge-in rate: Percentage of agent responses that were interrupted by the caller. A high rate may indicate the agent is talking too much or too slowly.
  • Duration distribution: Histogram of call durations. Helps identify whether calls are clustering around expected lengths or if many are ending prematurely.

Reliability​

  • Provider errors: Count and rate of errors by provider. Includes API timeouts, authentication failures, and rate limits.
  • Call failure rate: Percentage of calls that ended with a failed status.
  • Error breakdown: Table showing error types, affected providers, and timestamps.

Timing​

  • Peak hours: Heatmap showing call volume by hour and day of week. Useful for staffing decisions and capacity planning.
  • Average handle time: Mean call duration, excluding calls under 10 seconds (likely misdials or test calls).

Agent Performance​

  • Agent comparison: Side-by-side table of agents showing call volume, average duration, latency, cost per call, and completion rate.
  • Outcomes by agent: If you track call outcomes (e.g., sale made, appointment booked, issue resolved), this chart shows outcome distribution per agent.

Filtering​

Date Range​

Use the time filter at the top of the dashboard to select a time range:

  • Presets: Last Hour, Today, 7 Days, 30 Days, 90 Days.
  • Custom Range: Pick start and end dates.
  • All metrics and charts update to reflect the selected range.

Agent Filter​

Filter the entire dashboard to show metrics for a specific agent or a group of agents. This is useful when you want to compare how different system prompts or provider configurations perform.

Exporting Data​

Dashboard metrics are available via the VOCALS API for integration with external tools.

Real-Time Updates​

The dashboard updates automatically every 30 seconds while you have it open. Active calls are reflected in the call volume counter in near real-time. You can also click the refresh button for an immediate update.