PRISM D1 Velocity Assessment

AI-Assisted Development Lifecycle Maturity Report
Customer
Arcline Health
Team Size
14 engineers
Funding Stage
Series B
Assessment Date
2026-04-08
Solutions Architect
Priya Chakraborty
Repository
github.com/arcline-health/care-platform

Executive Summary

L2.5
PRISM D1 Level 2.5
Structured — AI tooling adopted, early measurement
Ready For Pilot

Arcline Health is a Series B startup with 14 engineers, currently assessed at PRISM D1 Level 2.5 (Structured). The automated scanner scored 48/100 and the SA interview scored 54/100, with an org readiness score of 20/20, producing a blended score of 60.8. The assessment verdict is Ready For Pilot, routing Arcline Health to Track B: Full Workshop.

48
/ 100
Scanner Score (40%)
54
/ 100
Interview Score (40%)
20
/ 20
Org Readiness (20%)

Scanner Category Breakdown

AI Tool Config Spec-Driven .. Commit Hygiene CI/CD Integr.. Eval & Quality Testing Matu.. AI Observabi.. Governance Agent Workfl.. Platform Reuse Documentation Dependencies
CategoryScoreProgressStatus
AI Tool Config 8 / 10
80%
Spec-Driven Dev 5 / 10
50%
Commit Hygiene 10 / 15
66%
CI/CD Integration 7 / 15
46%
Eval & Quality 2 / 10
20%
Testing Maturity 6 / 10
60%
AI Observability 1 / 10
10%
Governance 2 / 5
40%
Agent Workflows 1 / 5
20%
Platform Reuse 2 / 5
40%
Documentation 2 / 3
66%
Dependencies 2 / 2
100%

Interview Assessment

SectionScoreStatus
AI Tooling Landscape
  • Claude Code deployed via Bedrock across all 3 squads
  • CLAUDE.md exists but is inconsistent across repos (2 of 5 repos have it)
  • Kiro IDE adopted by 8 of 14 developers
  • Standardized on Claude 3.5 Sonnet for code generation
 11 / 15 73%
Development Workflow & Specs
  • Feature specs used informally by senior engineers but not mandated
  • No bug fix or refactor spec templates in use
  • PRD-to-spec mapping attempted but not systematic
  • Team lead has drafted a spec template but it is not adopted yet
 10 / 20 50%
CI/CD & Quality
  • GitHub Actions with lint, test, and build steps
  • No eval gates -- all PRs go through human review only
  • Test coverage at 68%, above average for the stage
  • HIPAA compliance checks in CI are manual
 10 / 20 50%
Metrics & Visibility
  • Basic GitHub Insights used for commit activity
  • No AI-specific metrics (acceptance rate, cost, token usage)
  • Engineering manager tracks velocity in Jira but not AI contribution
  • No executive dashboard or readout process
 4 / 15 26%
Governance & Security
  • HIPAA-aware: team avoids sending PHI to AI tools
  • No formal AI governance policy documented
  • Data handling guidelines exist for prod but not for AI tools
  • Security team has reviewed Bedrock data residency
 7 / 15 46%
Org & Culture
  • VP Engineering is the executive sponsor, actively engaged
  • Dedicated AI champion (senior engineer) driving adoption
  • All 3 squads have at least one AI-enthusiastic member
  • Team ran an internal AI hackathon last quarter
 12 / 15 80%

Organizational Readiness

✓Executive Sponsor Identified
✓Dedicated AI Champion
✓Team Willingness
✓Budget Approved
✓Timeline Commitment

Org readiness score: 20/20

Top Strengths

  1. Dependencies: 2/2 (100%)
  2. AI Tool Config: 8/10 (80%)
  3. Commit Hygiene: 10/15 (66%)

Gap Analysis & Remediation

PriorityCategoryScoreRecommended Action
#1 AI Observability (scanner) 1/10 (10%) Deploy the EventBridge-Timestream metrics pipeline. Enable token tracking, cost attribution, and latency monitoring.
#2 Eval & Quality (scanner) 2/10 (20%) Define quality rubrics for AI-generated code. Implement automated scoring. Establish minimum acceptance thresholds.
#3 Agent Workflows (scanner) 1/5 (20%) Identify first candidate for multi-step agent workflow. Start with low-risk automation. Document the workflow pattern.

Onboarding Recommendation

Track B: Full Workshop

Workshop Modules

IDModuleRationale
✓ M00 Environment Setup Verify and upgrade existing setup
✓ M01 CLAUDE.md & Standards Align existing config to PRISM standards
✓ M02 Spec-Driven Development Formalize and standardize spec workflows
✓ M03 CI/CD & Eval Gates Instrument pipeline with eval gates
✓ M04 Metrics & Dashboards Deploy Timestream + QuickSight dashboards
✓ M05 Governance & Scaling Establish governance model for scaling

Success Metrics

MetricTargetMeasure By
AI acceptance rate 30%+ Week 4 checkpoint
Eval gate active in CI At least 1 pipeline Week 2 of pilot
Weekly executive readout Active and reviewed Week 3 of pilot
PRISM D1 level improvement +0.5 levels Week 8 of pilot
Dashboard data freshness Within 1 hour Week 2 of pilot

90-Day Roadmap

TimelineMilestoneMeasurable Outcome
Week 1 Bootstrapper fully deployed All components live, first metrics flowing
Week 2 Eval gate in CI, dashboards active PR pipeline includes eval step, QuickSight live
Week 4 Midpoint checkpoint passed AI acceptance rate 30%+, exec readout active
Week 6 Governance model documented Policy document published to team wiki
Week 8 Pilot readout delivered L3.0+ achieved, Track C eligible
Week 12 90-day sustained adoption Metrics stable, no regression below L2.5

SA Engagement Cadence

WeekTypeDurationAgenda
Week -1 Pre-work kick-off call 30 min Review pre-work checklist, confirm AWS access, workshop logistics
Week 0 Workshop delivery 4 hr Deliver all 6 modules, hands-on exercises, bootstrapper deployment
Week 1 Video call 45 min Verify bootstrapper deployment, review first metrics, troubleshoot
Week 2 Async check-in 15 min (Slack) Review dashboard data quality, flag any pipeline issues
Week 4 Checkpoint call 1 hr Midpoint review, acceptance rate check, adjust pilot targets
Week 6 Async check-in 15 min (Slack) Review progress toward 8-week goals, flag risks
Week 8 Pilot readout call 1 hr Final pilot metrics, L3 readiness determination, next steps

PRISM D1 Velocity Assessment Report — Generated 2026-04-08

AWS Solutions Architecture · Startups Organization · CONFIDENTIAL

Solutions Architect: Priya Chakraborty