prism-d1-velocity

PRISM D1 Velocity – Executive Readout

Prepared for: [CTO / VP Engineering Name] Prepared by: [SA Name], AWS Solutions Architect Period: [Week/Sprint of YYYY-MM-DD to YYYY-MM-DD] Customer: [Company Name]

1. PRISM Level Summary

	Current	Target	Trend
PRISM Level	L[X] [Name]	L[X+1] [Name]	[Up/Flat/Down]
Assessed Date	YYYY-MM-DD	Target: YYYY-MM-DD

Level Definition Reminder:

L1 Ad Hoc – No systematic AI integration
L2 Emerging – Individual AI tool adoption, basic metrics
L3 Scaling – Team-wide AI workflows, eval gates in place
L4 Optimized – AI-native development, continuous evaluation
L5 Transformative – Autonomous AI pipelines, self-improving systems

Key blockers to next level:

[e.g., Eval gate coverage below 80% threshold for L3]
[e.g., AI acceptance rate inconsistent across teams]
[e.g., Spec-to-code pipeline not yet instrumented]

2. Enhanced DORA Scorecard

Traditional DORA Metrics

Metric	Current Value	L[X] Target	L[X+1] Target	Status	WoW Change
Deployment Frequency	[X] deploys/day	[target]	[target]	[Green/Amber/Red]	[+/-X%]
Lead Time for Changes	[X] hours	[target]	[target]	[Green/Amber/Red]	[+/-X%]
Change Failure Rate	[X]%	[target]%	[target]%	[Green/Amber/Red]	[+/-X pp]
MTTR	[X] hours	[target]	[target]	[Green/Amber/Red]	[+/-X%]

AI-DORA Extension Metrics

Metric	Current Value	L[X] Target	L[X+1] Target	Status	WoW Change
AI Acceptance Rate	[X]%	[target]%	[target]%	[Green/Amber/Red]	[+/-X pp]
AI-to-Merge Ratio	[X.XX]	[target]	[target]	[Green/Amber/Red]	[+/-X]
Eval Gate Pass Rate	[X]%	[target]%	[target]%	[Green/Amber/Red]	[+/-X pp]
Spec-to-Code Turnaround	[X] hours	[target]h	[target]h	[Green/Amber/Red]	[+/-X%]
Post-Merge Defect Rate (AI)	[X.XX]	[target]	[target]	[Green/Amber/Red]	[+/-X]
AI Test Coverage Delta	[+/-X] pp	[target] pp	[target] pp	[Green/Amber/Red]	[+/-X pp]

Status legend: Green = at or exceeding target for current level. Amber = within 20% of target. Red = more than 20% below target.

3. AI Adoption Highlights

What is working

[e.g., Claude Code adoption reached 45% acceptance rate on the Platform team, up from 32% two weeks ago]
[e.g., AI-generated tests are achieving 12% higher branch coverage than human-written tests]
[e.g., Spec-to-code turnaround dropped from 36h to 18h after Kiro integration]

What needs attention

[e.g., Payments team AI acceptance rate declined to 22% – investigating prompt quality issues]
[e.g., Eval gate pass rate dropped to 71% after model upgrade – retuning eval suite]
[e.g., Q Developer generating high false-positive rate on security suggestions]

Tool-specific observations

Tool	Teams Using	Acceptance Rate	Notable
Claude Code	[X] teams	[X]%	[observation]
Amazon Q Developer	[X] teams	[X]%	[observation]
Kiro	[X] teams	[X]%	[observation]

4. Investment vs. Return

Spend This Period

Category	Amount	Change
Amazon Bedrock (Claude)	$[X,XXX]	[+/-X%]
Amazon Q Developer licenses	$[X,XXX]	[flat]
Kiro (if applicable)	$[X,XXX]	[+/-X%]
Timestream + observability	$[XXX]	[flat]
Total AI Platform Spend	$[X,XXX]	[+/-X%]

Value Delivered This Period

Dimension	Estimated Value	Calculation
Velocity: Engineering time saved	$[XX,XXX]	[X] hours saved x $[XXX]/hr fully-burdened
Quality: Defects prevented	$[X,XXX]	[X] fewer defects x $[X,XXX] avg incident cost
Eval: Incidents prevented	$[X,XXX]	[X] eval-blocked merges x $[X,XXX] est. blast radius
Total Estimated Value	$[XX,XXX]
ROI Multiplier	[X.X]x	Value / Spend

See ROI Model for methodology and sensitivity analysis.

5. Recommendations

Immediate (This Sprint)

[Action] – [Rationale]. Owner: [Name]. Expected impact: [metric improvement].
[Action] – [Rationale]. Owner: [Name]. Expected impact: [metric improvement].

Near-term (Next 2-4 Weeks)

[Action] – Required for L[X+1] advancement. Dependency: [what].
[Action] – Will unlock [capability]. Requires: [resource/decision].

Strategic (Next Quarter)

[Action] – Positions team for L[X+2]. Investment required: [estimate].

6. Risk Register

Risk	Likelihood	Impact	Mitigation	Owner
[e.g., Eval gate coverage falls below L3 threshold]	[High/Med/Low]	[High/Med/Low]	[action]	[name]
[e.g., Model version change degrades acceptance rate]	[Med]	[Med]	[Pin model versions, A/B test upgrades]	[name]
[e.g., Team resistance to AI-generated code review]	[Med]	[High]	[Champion program, show defect data]	[name]
[e.g., Bedrock cost scaling faster than value]	[Low]	[High]	[Usage tiers, prompt optimization]	[name]
[e.g., Data pipeline latency masks real-time issues]	[Low]	[Med]	[CloudWatch alarms, Timestream TTL]	[name]

Appendix: Dashboard Links

Team Velocity (CloudWatch): https://REGION.console.aws.amazon.com/cloudwatch/home?region=REGION#dashboards:name=PRISM-D1-TeamVelocity
Executive Readout (CloudWatch): https://REGION.console.aws.amazon.com/cloudwatch/home?region=REGION#dashboards:name=PRISM-D1-ExecutiveReadout
AI-DORA Analysis (QuickSight): https://REGION.quicksight.aws.amazon.com/sn/analyses/prism-d1-ai-dora-analysis
Level Tracker (QuickSight): https://REGION.quicksight.aws.amazon.com/sn/analyses/prism-d1-level-tracker

This readout was generated using the PRISM D1 Velocity metrics platform. Data sourced from CloudWatch (PRISM/D1 namespace) and Timestream (prism_d1_velocity database).

Formatting guidance: Use Green/Amber/Red text or cell background in the final version (Google Slides, PowerPoint, or Notion). Replace all bracketed placeholders with actual values from the dashboards.