WebRTC Quality Analytics

Documentation / Screens

Screens

Screens are organized in a sidebar — Monitor (Overview, Sessions), Analyze (Breakdown, Compare, Call timeline), and Data (Upload dump). Global filters apply to all fleet views.

Global filters

Filters apply to Overview, Sessions, and Breakdown. Compare, Call timeline, and Upload are per-session views.

Overview

Overview screen

Fleet summary cards, daily MOS trend (average and 10th percentile), sessions per day, and QoE grade distribution.

The Last 7 days vs prior 7 days strip is an automatic regression check: deteriorations show red, improvements green. It uses a fixed 7-day window independent of the Period filter (dimension filters still apply).

Sessions

Sessions screen

Every analyzed call with MOS, loss, jitter, RTT, degradation ratio, TURN usage, and grade. Poor rows are tinted red.

Sort by any column, filter inline (e.g. type >100 in the RTT column), and click a row to open its call timeline.

Breakdown

Breakdown screen

Box plots per dimension with per-session points, grade counts, and P50/P95 distribution histograms.

Compare

Compare screen

Two sessions side by side: metric deltas in a table, full timelines below.

Pick a poor call and a good one from the same site — the degraded intervals light up on one side and not the other, and that difference is usually your answer.

Call timeline

Call timeline screen

Bitrate, packet loss, jitter/RTT, and E-model MOS at one-second resolution.

AI diagnosis

AI diagnosis panel

One click asks a foundation model in your workspace for an incident report that cites the evidence.

The diagnosis includes a severity badge (none / minor / major / critical), a plain-language summary, evidence with concrete numbers, likely root causes, and recommended actions. Every diagnosis is stored as history in gold_session_diagnosis next to your data. The prompt instructs the model to never invent metrics and to say so when a call was healthy.

Show SQL

Every fleet view has a "Show SQL for this view" panel with the exact, fully qualified query the app ran — paste it into the Databricks SQL editor, a notebook, or an AI/BI dashboard to go beyond the app.