L2 Decision Center#

Path: /decision

The decision center builds on L1 assessment results to analyze AI application scenarios and make investment decisions.

Overview#

Path: /decision/overview

The home dashboard shows: identified scenario count and distribution, ROI analysis overview, risk heatmap, portfolio progress.

Business Scenario Management#

Path: /decision/scenarios

Business scenarios are the core L2 entity; each scenario represents an AI application opportunity.

Scenario execution states:

State	Description
pending	Pending evaluation
selected	Selected for analysis
scheduled	Scheduled
in_delivery	In delivery (enters L4)
completed	Completed
cancelled	Cancelled

Analysis dimensions:

• ROI analysis: investment return based on cost/benefit
• Risk assessment: implementation risks (schedule/budget/scope/technical/resource/stakeholder/external)
• Prioritization: multi-dimension weighted scoring

Priority scoring weights:

Dimension	Weight
Business value (business_value)	25%
Strategic alignment (strategic_alignment)	15%
Data readiness (data_readiness)	15%
Technical feasibility (technical_feasibility)	15%
Implementation risk (implementation_risk)	15%
ROI	15%

Priority thresholds:

Level	Score range
P0	≥ 85
P1	≥ 70
P2	≥ 50

Steps:

1. Click New Scenario to create
2. Fill in basic info (business domain, AI Pattern, expected effect)
3. The system automatically computes ROI and risk assessment
4. Use AI Analysis for a detailed AI-generated analysis report
5. Add the scenario to a portfolio
6. Submit for approval

AI Pattern types:

Pattern	Dependencies	Description
RAG	—	Retrieval-augmented generation
AGENT	RAG	Intelligent agent
WORKFLOW	—	Workflow automation
LLM	—	Large language model application
ML	—	Machine learning
HYBRID	RAG + ML	Hybrid mode

💡 AGENT requires RAG data preparation first; HYBRID requires both RAG and ML foundations.

Decision Analysis#

Path: /decision/analysis

ROI analysis (ROI tab) — computed from:

Parameter	Values
Complexity base cost	low: ¥150K / medium: ¥400K / high: ¥800K
Enterprise scale factor	startup: 0.6 / SME: 0.8 / mid: 1.0 / large: 1.3 / enterprise: 1.6
Maturity factor	initial: 1.5 / developing: 1.2 / defined: 1.0 / managed: 0.8 / optimizing: 0.7
Benefit multiplier	low: 1.5x / medium: 2.5x / high: 4.0x

Risk assessment (Risk tab) — risk category to dimension mapping:

Risk category	Related dimensions
data	Data
organizational	Talent, Culture
implementation	Technology, Talent
technical	Technology, Data
compliance	Governance, Strategy

Prioritization (Priority tab): the system sorts automatically by the 6-dimension weighted score, outputting P0/P1/P2.

Solution comparison (/decision/solution-compare): side-by-side comparison of multiple scenarios/AI Patterns.

Solution Pattern management (/decision/solution-patterns): manage predefined AI solution patterns.

Portfolio Optimization#

Path: /decision/portfolio

Portfolios help optimize AI investment allocation.

Portfolio stages:

Stage	Time range	Description
Quick win (quick_win)	0-6 months	Fast-payoff AI projects
Core transformation (core)	6-12 months	Core business AI transformation
Strategic (strategic)	12-24 months	Long-term strategic AI investment

Steps:

1. View the three-stage view on the portfolio page
2. Drag selected scenarios into stages
3. The system computes per-stage budget and resource needs automatically
4. Adjust scenario priorities and scheduling
5. Generate the portfolio report

Decision Approval#

Path: /decision/review

Decision approval is the final approval flow for business scenarios.

Steps:

1. View scenarios awaiting approval
2. Review the full analysis report (ROI, risks, priority)
3. Choose an action:
   • Approve: the scenario moves to the next stage
   • Conditionally approve: pass with conditions
   • Request revision: send back for revision
   • Reject: veto the scenario
4. Submit your approval comment

Decision report (/decision/reports): after approval, generate the decision report — the input for L3 Planning.