Support Dashboard
Support Dashboard
Executive Summary
Release Intelligence · GitHub Analysis
Verdict
Bug volume has declined 31% since the January peak — from 68 bugs in Jan 5–18 down to 47 in Feb 16–27 — driven by the adoption of AI-assisted planning, execution checks, and PR guardrails that introduced structural discipline into how features are scoped, reviewed, and shipped.
The Jan 5–18 spike was release-driven: 5+ features hit production simultaneously, producing the all-time bug peak of 68. All 4 periods had a release component — including the Position Backdating core module landing in Feb 2–15 — yet bugs continued to fall, confirming that engineering quality controls are now absorbing release pressure that previously translated directly into support volume.
68
Jan 5–18
Bugs at peak
47
Feb 16–27
Bugs now
−31%
Bug reduction
since peak
Backend + Frontend combined releases per period overlaid on issue count
Total merged PRs (Backend + Frontend) — proxy for engineering change velocity
281
merged PRs
Major features shipped to staging:
Backend: 164 releases · Frontend: 117 releases.
316
merged PRs
Staging features promoted to PRODUCTION simultaneously:
Frontend: 40 releases in 9 days (Jan 5–13). Backend: 0 formal releases (CI/CD direct to prod).
331
merged PRs
Jan 5–18 launch fallout and active backdating development:
Highest PR count of any period (331) — engineering simultaneously firefighting Jan launches while building Position Backdating in parallel.
323
merged PRs
Position Backdating core module ships + ongoing Jan complexity:
323 PRs merged. The Backdating core module (9 PRs in 5 days) landing simultaneously with high outstanding issue volume made this the second-highest spike period.
⚠ Release Bundling Drove the Bug Peak
5 major features (Custom Sales Fields, Worklio, SequiPay, Everee Webhooks, Sales Pusher) hit production simultaneously in Jan 5–18, driving 68 bugs — the all-time high. Bundling complex releases made root-cause isolation difficult and compressed the team's ability to respond.
▲ Bug Reduction in Action
Bugs per period: 68 → 52 → 44 → 47. A 31% reduction since the January peak — even as the team continued shipping major features at pace. The slight uptick to 47 in Feb 16–27 is within normal variance and does not break the downward trend.
🔧 High Velocity on an Unstable Base
Jan 19–Feb 15 saw the highest PR volumes of the entire dataset (331 + 323 PRs). Rather than pausing to stabilise, the team continued shipping new features — Custom Sales Fields, Position Backdating, Sales Process refactors — at full speed onto a system already carrying a significant open issue load. New complexity was being layered onto unresolved complexity, amplifying each subsequent spike.
📅 Position Backdating Timeline
Backdating was a 5-week rolling release: inception Jan 16, active UI/BE dev Jan 19–Feb 1, core module to prod Feb 10–14, custom fields persistence Feb 16–17. Spanning 3 consecutive spike periods, it is the single longest-running feature thread in this dataset.
Overview
Section 1 — Severity
Stacked by severity across bi-weekly periods
Tracks highest-severity issues per period — key risk indicator
Section 2 — Engineering Throughput
% of tickets with Status = Done within each period
Count of non-Done tickets by status category
Section 3 — Direction of Travel
Actual bug count per bi-weekly period with trendline — all other issue types excluded
Bug % change vs previous period increase decrease
% of bugs that are Critical or High priority each period — are bugs getting more serious?
Each functional area shown separately across all 4 periods — P3 vs P2 trend indicator shown
Section 4 — Critical Load Concentration
Top 10 clients — which clients carry the most critical load
Top 10 functional areas — which areas attract the highest-severity issues
Section 5 — Product Engineering Focus
Top 9 issue labels + Other as doughnut
Stacked bar — which areas were most affected each period
Section 6 — Customer Impact
Horizontal bar sorted by total issue count (excl. "All")
Stacked bar — which clients drove volume each period
Section 7 — Nature of Problems
Stacked count by issue type across all periods
Count of each type per period — spot shifts in the mix
100% stacked — share of bugs vs all other types each period
All-time breakdown across 370 filtered issues
Section 8 — Volume Context
Bi-weekly issue volume P4 = partial / current period
Running total — shows overall pace of issue reporting
Section 9 — Operational Detail
Repository split across periods
Breadth of impact — distinct clients raising issues
Issue count by priority (row) and type (col) — darker = higher