Demo

Arcline Health Engineering

AI-DLC Executive Readout — Week of Apr 7, 2026 · Sprint 14

PRISM Level L2.5 Structured+
6 Teams
Onboarded
42 Active
Engineers

AI Tool Adoption & Licensing

Enabled vs. actively using — identify adoption gaps and unused licenses across teams.

42/48
Licenses Active
88% utilization
38
Weekly Active Users
+6 vs last month
6
Unused Licenses
$2.4K/mo potential savings
14days
Avg Time to First Use
-5 days vs last cohort

Adoption by Tool

Current
Claude Code 36/42 86% active Kiro IDE 28/42 67% active Bedrock Evals 22/42 52% active AI Metric Hooks 42/42

Adoption Trend

12 Weeks
0 25 42 AI-DLC Claude Code Kiro IDE

Speed Impact

PR cycle time and throughput — comparing teams with different AI adoption levels.

2.1hrs
Median PR Cycle Time
-34% vs pre-AI baseline
4.2/day
Deploy Frequency
+18% vs last sprint
187
PRs Merged This Sprint
+24 vs last sprint
1.8hrs
Spec-to-Code Turnaround
-57% vs baseline

PR Cycle Time: AI-Assisted vs Human-Only

Last 30 Days
0 8 hrs 4 hrs Coding 5.2h 1.4h Review 4.0h 2.0h Deploy 1.8h 1.2h Total 7.5h 2.1h Human-Only PRs AI-Assisted PRs -72% faster

Throughput by Team (PRs/week)

Current Sprint
Platform AI: 71% 34 Backend API AI: 67% 31 Data Pipeline AI: 61% 27 Frontend AI: 57% 24 Mobile AI: 49% 19 QA/DevOps AI: 52% 21 AI-Assisted Human

Quality & Batch Size

Monitoring PR sizes, change failure rates, and eval gate effectiveness to catch AI-generated defects early.

3.2%
Change Failure Rate
-1.1pp vs baseline
148LOC
Median PR Size
+22% — watch batch size
91%
Eval Gate Pass Rate
+4pp vs last sprint
28min
Mean Time to Recovery
-12min vs last sprint

PR Size Distribution

AI vs Human PRs
<50 LOC 50-150 150-300 300-500 500+ Review carefully Human PRs AI-Assisted PRs

Eval Gate Failures by Category

Last 30 Days · 17 blocked
Hallucinated APIs 8 Spec Mismatch 4 Missing Error Handling 3 Logic Errors 2 17 defects caught before merge — $0 production impact

Investment Balance & ROI

Where engineering time goes and the financial return on AI-DLC investment.

Engineering Investment Balance

This Quarter
% of eng time Pre-AI 40% maint 60% features Post-AI (Now) 24% maint 76% features +16pp

Quarterly ROI Summary

Q2 2026
$284K
Eng Hours Saved
~4,730 hrs @ $60/hr avg
$18K
Total AI Spend
Bedrock + tool licenses
15.8x
Return on AI Investment
$284K saved / $18K spent
62%
Fewer Post-Merge Defects
1.4% vs 3.7% baseline

Agent-Created PRs

Coding agents (autonomous)
23
Agent PRs Created
18
Merged
5
Closed (not merged)
78%
Agent Merge Rate
12%
Share of All PRs