Skip to main content

Insights Summary

Overview​

The Summary tab gives you a 30-day rollup of Aurora Insights activity — total findings, breakdown by priority, daily trend, top-firing rules, and which agents have the most flagged conversations. Use it to spot patterns over time: which detection rules are working, which agents need coaching, and whether the volume of issues is trending up or down.

How to Access​

  • Click Insights in the left sidebar.
  • Open the Summary tab.
  • Visible only to admins.

Interface Overview​

The Summary page has three sections, top to bottom:

  1. Stats strip — four big-number tiles:
    • Total Findings — sum of all findings in the last 30 days.
    • Critical / Important / Info — counts by priority.
  2. Daily Trend (Last 30 days) — a stacked bar chart with one bar per day, color-coded by priority. If your account has only a handful of days of data, a sparse-data message appears in place of the chart.
  3. Split breakdown — two side-by-side cards:
    • Findings by Rule — every rule that produced findings in the window, ranked by count, with a horizontal bar showing relative volume.
    • Findings by User — table of agents with columns for Critical, Important, Info, and Total. Empty cells render as a dash.

Features & Actions​

Read the Daily Trend​

What it does: Shows day-by-day finding counts, stacked by priority, for the last 30 days.

How to use it:

  • Each bar is one day. The colored segments stack Critical (top), then Important, then Info.
  • Hover any bar to see the exact count for each priority that day.
  • Use this view to spot weekly patterns (Mondays often have higher counts after weekend backlogs), incident days (one-time spikes), and longer-term trends after you launch new rules or coach your team.

Identify the Most-Fired Rule​

What it does: Shows which detection rules produced the most findings.

How to use it:

  • The Findings by Rule card lists every rule with at least one finding in the window.
  • Each row shows the rule name, the count, and a horizontal bar.
  • The bar uses a logarithmic scale so a rule with two findings still shows a visible bar next to a rule with two hundred — small counts are not crushed to invisibility.

Spot Workload Imbalance Across Agents​

What it does: Shows which agents have the most flagged conversations.

How to use it:

  • The Findings by User table lists every agent who appears as the owner on at least one finding.
  • Columns split the count by priority so you can tell whether an agent has a lot of low-priority findings (volume issue) versus a handful of Criticals (coaching opportunity).
  • Sort visually by scanning the Total column.

Fields Reference​

FieldDescription
Total FindingsSum of all findings in the last 30 days, regardless of status
CriticalNumber of findings with Critical priority
ImportantNumber of findings with Important priority
InfoNumber of findings with Info priority
Daily TrendOne bar per day for the last 30 days, segmented by priority
Findings by Rule (count)Number of findings produced by each rule in the window
Findings by Rule (bar)Visual share of total volume, log-scaled to keep small values visible
Findings by User (Critical / Important / Info)Per-priority count of findings owned by each agent
Findings by User (Total)Sum across priorities for each agent

Business Rules​

  • The Summary always shows the last 30 days. The window is not user-configurable on this page.
  • Findings of every status are counted — New, Reviewed, Dismissed, Action Taken, and Auto-Resolved. The Summary measures detection volume, not unresolved-issue volume.
  • The Findings by Rule bar is logarithmic by design so that low-volume rules (and brand-new custom rules) are still visible next to high-volume ones.
  • The Findings by User table is populated from the Owner field on each finding — if a conversation has no agent assigned at the time the finding is produced, it does not appear in the per-user breakdown.
  • If your account has fewer than a few days of data, the trend chart is replaced by a "X days" placeholder. The byRule and byUser cards still render normally.
  • This page is visible to admins only because it includes per-agent breakdowns.

FAQ​

Q: Why is the bar chart so flat for our top rule? Other rules have tiny counts but their bars look almost the same size. A: The Findings by Rule chart uses a logarithmic scale on purpose. With a linear scale, a rule with two findings would be invisible next to a rule with two hundred. Log scale lets you see all your rules at a glance — but it does mean the bars are not directly proportional to count.

Q: Why doesn't a specific agent appear in the byUser table? A: The byUser table is grouped by the Owner of each conversation at the time the finding was produced. If the conversation had no owner, or the owner is no longer an active user, they may not appear. Findings on team-shared conversations without a specific owner are not counted per-user.

Q: Can I export this to CSV or share a link? A: Not from the Summary tab today. For per-finding data, use the Dashboard tab and filter by the date range you care about.

Q: Can I change the window to last 7 days or last 90 days? A: Not from this page. The 30-day window is fixed. For finer-grained slices, use the Dashboard's date-range filter.

Q: Total Findings on the Summary doesn't match what I see on the Dashboard. Why? A: The Summary counts every finding produced in the last 30 days regardless of status. The Dashboard's quick-pill strip only counts New findings by default. Apply a "All statuses" filter on the Dashboard and a 30-day date range to compare like-for-like.