Engineering Flow
AI-first executive view · Custom scope · last 30d
AI Bottom Line
Delivery slowed this period: throughput fell to 75 issues done while median Jira flow lead time remained at 623 hours, with the biggest drag in waiting states before work is ready and while work waits on external response.
Headline KPIs · last 30d
AI insight
- Where AI helps:AI-owned work shows materially lower median lead time at 169.0 hours and lower median cycle time at 147.2 hours; AI-touched work is also lower at 237.8 hours lead time and 153.3 hours cycle time.
- Where AI hurts:AI-owned work still accumulates waiting in Ready for Staging, and overall team flow remains dominated by non-AI bottlenecks such as Waiting for Client and On Hold.
- Next focus:Attack waiting-state aging first: review items in Waiting for Client, On Hold, and long-stuck in-progress work, then tighten staging handoff before adding more intake.
AI in teams · 1 team
75 completed · 40.0% AI-touched · 1 active agent
Bottom line
EVO declined (-47.2%) in throughput. AI-owned work moves faster on cycle (n=23/48); the main AI bottleneck is Ready For Staging.
Issues done
75
-47.2% vs prev
AI-owned share
36%
+17.0 pp vs prev
AI-touched
30 (40%)
Lead Time (active)
26d
prev 18.2d
Cycle Time (active)
6.7d
prev 6.1d
TTR p50
13.7d
prev 6.9d
PR Review p50
—
WIP active
42
WIP blocked
5
AI insight
Where AI helps
- · AI Cycle Time (active) 6.1d vs Human 7d (n=23/48)
- · AI Lead Time (active) 7d vs Human 33.5d (n=27/48)
- · AI Time to Ready p50 1.8d vs Human 16.9d (n=23/48)
Where AI hurts
No AI-owned metric is currently slower than the human cohort.
Top AI bottleneck
Active agents · top 1
AI Period: compare · intro 2026-04-02
Active bottlenecks · top 5
Top teams · by throughput
| Key | Team | Throughput | Lead Time (active) | Cycle Time (active) | Active WIP | Blocked WIP | AI % |
|---|---|---|---|---|---|---|---|
| EVO | Platforms Evolution | 75 | 26d | 6.7d | 42 | 5 | 36% |