# 🤖 Agentic AI Readiness Assessment

**Version:** 1.0
**Date:** ___________________________
**Organization:** ___________________________
**Assessor:** ___________________________
**Current Maturity Level:** ___/6

---

## 📋 Assessment Overview

Agentic AI represents the next evolution of GenAI, where AI systems can autonomously plan, execute multi-step tasks, use tools, and make decisions with varying degrees of human oversight. This assessment evaluates your organization's readiness to deploy agentic AI capabilities.

**Important:** Agentic AI requires a strong foundation in traditional GenAI. Organizations should typically be at **maturity level 4 or higher** before pursuing agentic deployments.

---

## ✅ Section 1: Foundation Prerequisites

### 1.1 GenAI Maturity Baseline

_Confirm the following foundational elements are in place:_

| Prerequisite | Status | Evidence | Gap |
|--------------|--------|----------|-----|
| **GenAI Success Track Record** | | | |
| - 3+ GenAI use cases in production | ⬜ Yes ⬜ Partial ⬜ No | _________ | _________ |
| - Demonstrated ROI from GenAI | ⬜ Yes ⬜ Partial ⬜ No | _________ | _________ |
| - User adoption >50% for deployed tools | ⬜ Yes ⬜ Partial ⬜ No | _________ | _________ |
| **Governance Foundation** | | | |
| - AI governance framework in place | ⬜ Yes ⬜ Partial ⬜ No | _________ | _________ |
| - AI ethics board/committee active | ⬜ Yes ⬜ Partial ⬜ No | _________ | _________ |
| - Risk assessment process established | ⬜ Yes ⬜ Partial ⬜ No | _________ | _________ |
| **Oversight Capabilities** | | | |
| - AI monitoring in production | ⬜ Yes ⬜ Partial ⬜ No | _________ | _________ |
| - Human-in-the-loop processes defined | ⬜ Yes ⬜ Partial ⬜ No | _________ | _________ |
| - Model performance tracking | ⬜ Yes ⬜ Partial ⬜ No | _________ | _________ |
| **Incident Response** | | | |
| - AI incident response plan exists | ⬜ Yes ⬜ Partial ⬜ No | _________ | _________ |
| - Incident response tested | ⬜ Yes ⬜ Partial ⬜ No | _________ | _________ |
| - Rollback procedures documented | ⬜ Yes ⬜ Partial ⬜ No | _________ | _________ |

**Prerequisites Score:** ___/12 "Yes" responses

| Score | Readiness |
|-------|-----------|
| 10-12 | Ready to proceed with agentic assessment |
| 7-9 | Address gaps before proceeding |
| <7 | Focus on foundational GenAI maturity first |

---

## 🔧 Section 2: Technical Readiness

### 2.1 Orchestration Layer

| Capability | Score (1-5) | Current State | Gap |
|------------|-------------|---------------|-----|
| Workflow orchestration platform | ___/5 | _________________________ | _________ |
| Agent framework experience | ___/5 | _________________________ | _________ |
| State management capabilities | ___/5 | _________________________ | _________ |
| Error handling & recovery | ___/5 | _________________________ | _________ |
| Async/parallel execution | ___/5 | _________________________ | _________ |

**Section Score:** ___/25

### 2.2 Tool & API Integration

| Capability | Score (1-5) | Current State | Gap |
|------------|-------------|---------------|-----|
| Internal APIs available | ___/5 | _________________________ | _________ |
| API authentication/authorization | ___/5 | _________________________ | _________ |
| API rate limiting & management | ___/5 | _________________________ | _________ |
| External service integration | ___/5 | _________________________ | _________ |
| API documentation completeness | ___/5 | _________________________ | _________ |

**Section Score:** ___/25

### 2.3 Memory & Context Management

| Capability | Score (1-5) | Current State | Gap |
|------------|-------------|---------------|-----|
| Short-term memory (session) | ___/5 | _________________________ | _________ |
| Long-term memory (persistent) | ___/5 | _________________________ | _________ |
| Context window optimization | ___/5 | _________________________ | _________ |
| Knowledge retrieval (RAG) | ___/5 | _________________________ | _________ |
| Memory pruning/summarization | ___/5 | _________________________ | _________ |

**Section Score:** ___/25

### 2.4 Guardrails & Safety Systems

| Capability | Score (1-5) | Current State | Gap |
|------------|-------------|---------------|-----|
| Input validation/filtering | ___/5 | _________________________ | _________ |
| Output validation/filtering | ___/5 | _________________________ | _________ |
| Action confirmation gates | ___/5 | _________________________ | _________ |
| Scope/boundary enforcement | ___/5 | _________________________ | _________ |
| Fallback/degradation mechanisms | ___/5 | _________________________ | _________ |

**Section Score:** ___/25

### 2.5 Monitoring & Observability

| Capability | Score (1-5) | Current State | Gap |
|------------|-------------|---------------|-----|
| Real-time agent monitoring | ___/5 | _________________________ | _________ |
| Decision/action logging | ___/5 | _________________________ | _________ |
| Cost tracking per agent/task | ___/5 | _________________________ | _________ |
| Latency/performance metrics | ___/5 | _________________________ | _________ |
| Alerting for anomalies | ___/5 | _________________________ | _________ |

**Section Score:** ___/25

### 2.6 Technical Readiness Checklist

**Essential Infrastructure:**
- [ ] LLM API access (with appropriate rate limits)
- [ ] Vector database for knowledge retrieval
- [ ] Secure credential management
- [ ] Containerized/isolated execution environment
- [ ] Logging and audit trail infrastructure
- [ ] Development/staging/production environments

**Recommended Tools & Frameworks:**
- [ ] Agent framework (e.g., LangChain, AutoGen, CrewAI)
- [ ] Orchestration platform (e.g., Temporal, Airflow)
- [ ] Observability stack (e.g., LangSmith, Helicone)
- [ ] Testing framework for agents
- [ ] CI/CD pipeline for agent deployment

**Technical Readiness Total:** ___/125

---

## 🤝 Section 3: Multi-Agent Considerations

### 3.1 Multi-Agent Architecture Assessment

_If planning to deploy multiple interacting agents:_

| Consideration | Score (1-5) | Current Capability | Plan |
|---------------|-------------|-------------------|------|
| Agent-to-agent communication | ___/5 | _________________ | _________________ |
| Coordination mechanisms | ___/5 | _________________ | _________________ |
| Conflict resolution logic | ___/5 | _________________ | _________________ |
| Shared state management | ___/5 | _________________ | _________________ |
| Hierarchical oversight | ___/5 | _________________ | _________________ |

**Section Score:** ___/25

### 3.2 Multi-Agent Patterns

_Which patterns are you considering?_

- [ ] **Sequential Chain:** Agents process in order, output passes to next
- [ ] **Parallel Execution:** Multiple agents work simultaneously
- [ ] **Hierarchical:** Manager agent delegates to worker agents
- [ ] **Collaborative:** Agents discuss/debate to reach consensus
- [ ] **Competitive:** Agents compete, best output selected
- [ ] **Specialist Routing:** Router directs to domain-specific agents

### 3.3 Multi-Agent Risks

| Risk | Likelihood (1-5) | Impact (1-5) | Mitigation |
|------|------------------|--------------|------------|
| Agent loops/infinite recursion | ___ | ___ | _________________________ |
| Conflicting actions | ___ | ___ | _________________________ |
| Cost explosion | ___ | ___ | _________________________ |
| Cascading failures | ___ | ___ | _________________________ |
| Emergent unintended behavior | ___ | ___ | _________________________ |

---

## 🏢 Section 4: Organizational Readiness

### 4.1 Skills & Expertise

| Skill Area | Available | Gap | Plan |
|------------|-----------|-----|------|
| Agent architecture design | ⬜ Yes ⬜ Partial ⬜ No | _______ | _______ |
| Prompt engineering (advanced) | ⬜ Yes ⬜ Partial ⬜ No | _______ | _______ |
| API/integration development | ⬜ Yes ⬜ Partial ⬜ No | _______ | _______ |
| AI safety/alignment expertise | ⬜ Yes ⬜ Partial ⬜ No | _______ | _______ |
| System reliability engineering | ⬜ Yes ⬜ Partial ⬜ No | _______ | _______ |
| AI governance/compliance | ⬜ Yes ⬜ Partial ⬜ No | _______ | _______ |

### 4.2 Organizational Readiness Checklist

**Leadership & Culture:**
- [ ] Executive sponsorship for agentic AI initiatives
- [ ] Risk tolerance defined for autonomous actions
- [ ] Culture of experimentation and learning
- [ ] Clear escalation paths for agent issues

**Process Readiness:**
- [ ] Change management for AI-affected processes
- [ ] Training plan for staff working with agents
- [ ] Feedback loops for continuous improvement
- [ ] Documentation standards for agent behavior

**Stakeholder Alignment:**
- [ ] Legal review of autonomous action liability
- [ ] Compliance approval for agent capabilities
- [ ] Business process owners identified
- [ ] User acceptance criteria defined

---

## ⚖️ Section 5: Governance Enhancements for Agentic AI

### 5.1 Autonomy Level Framework

Define appropriate autonomy levels for different scenarios:

| Level | Description | Human Involvement | Use Cases |
|-------|-------------|-------------------|-----------|
| **Level 0** | Fully manual | Human does everything | Baseline |
| **Level 1** | AI assists | Human decides, AI helps | Research, drafting |
| **Level 2** | AI recommends | AI suggests, human approves | Analysis, recommendations |
| **Level 3** | AI acts, human monitors | AI executes, human reviews | Routine automation |
| **Level 4** | AI acts, human on exception | AI autonomous, human for edge cases | High-confidence tasks |
| **Level 5** | Fully autonomous | No human involvement | Narrow, well-defined tasks |

### 5.2 Autonomy Assignment Matrix

_Assign appropriate autonomy levels to potential agent tasks:_

| Task Category | Max Autonomy | Justification | Review Frequency |
|---------------|--------------|---------------|------------------|
| Information retrieval | Level ___ | _________________ | _________________ |
| Content generation | Level ___ | _________________ | _________________ |
| Data analysis | Level ___ | _________________ | _________________ |
| Internal communications | Level ___ | _________________ | _________________ |
| External communications | Level ___ | _________________ | _________________ |
| Financial transactions | Level ___ | _________________ | _________________ |
| System configurations | Level ___ | _________________ | _________________ |
| Customer interactions | Level ___ | _________________ | _________________ |
| HR/personnel actions | Level ___ | _________________ | _________________ |
| Compliance decisions | Level ___ | _________________ | _________________ |

### 5.3 Governance Enhancements Required

| Enhancement | Priority | Owner | Status |
|-------------|----------|-------|--------|
| Autonomous action policy | H/M/L | _______ | ⬜ Not Started ⬜ In Progress ⬜ Complete |
| Escalation triggers defined | H/M/L | _______ | ⬜ Not Started ⬜ In Progress ⬜ Complete |
| Agent audit trail requirements | H/M/L | _______ | ⬜ Not Started ⬜ In Progress ⬜ Complete |
| Liability framework updated | H/M/L | _______ | ⬜ Not Started ⬜ In Progress ⬜ Complete |
| Incident response updated for agents | H/M/L | _______ | ⬜ Not Started ⬜ In Progress ⬜ Complete |
| Agent testing/validation standards | H/M/L | _______ | ⬜ Not Started ⬜ In Progress ⬜ Complete |
| Cost controls for agent execution | H/M/L | _______ | ⬜ Not Started ⬜ In Progress ⬜ Complete |

---

## 📊 Section 6: Use Case Suitability Matrix

### 6.1 Suitability Assessment Framework

Evaluate potential agentic use cases on two dimensions:
- **Autonomy Benefit:** How much value does autonomous execution provide?
- **Risk Level:** What are the consequences of agent errors?

### 6.2 Use Case Evaluation

| Use Case | Autonomy Benefit (1-5) | Risk Level (1-5) | Suitability | Notes |
|----------|----------------------|------------------|-------------|-------|
| _____________________________ | ___ | ___ | ⬜ Suitable ⬜ With Controls ⬜ Not Suitable | _______ |
| _____________________________ | ___ | ___ | ⬜ Suitable ⬜ With Controls ⬜ Not Suitable | _______ |
| _____________________________ | ___ | ___ | ⬜ Suitable ⬜ With Controls ⬜ Not Suitable | _______ |
| _____________________________ | ___ | ___ | ⬜ Suitable ⬜ With Controls ⬜ Not Suitable | _______ |
| _____________________________ | ___ | ___ | ⬜ Suitable ⬜ With Controls ⬜ Not Suitable | _______ |

### 6.3 Suitability Quadrant

```
                    HIGH AUTONOMY BENEFIT
                           │
    ┌──────────────────────┼──────────────────────┐
    │                      │                      │
    │  PROCEED WITH        │     IDEAL FOR        │
    │  STRONG CONTROLS     │   AGENTIC AI         │
    │                      │                      │
    │  High benefit but    │   High benefit,      │
    │  high risk - needs   │   low risk -         │
    │  robust safeguards   │   prioritize these   │
    │                      │                      │
HIGH├──────────────────────┼──────────────────────┤LOW
RISK│                      │                      │RISK
    │  RECONSIDER          │     EVALUATE         │
    │  APPROACH            │   CAREFULLY          │
    │                      │                      │
    │  Low benefit, high   │   Low benefit, low   │
    │  risk - likely not   │   risk - may not     │
    │  worth pursuing      │   justify investment │
    │                      │                      │
    └──────────────────────┼──────────────────────┘
                           │
                    LOW AUTONOMY BENEFIT
```

### 6.4 Recommended Starting Use Cases

**Good First Agentic Use Cases (Low Risk, Clear Value):**
- [ ] Research and information gathering agents
- [ ] Document drafting and editing agents
- [ ] Data analysis and reporting agents
- [ ] Code review and documentation agents
- [ ] Meeting scheduling and coordination agents
- [ ] Internal knowledge Q&A agents

**Use Cases Requiring More Maturity:**
- [ ] Customer-facing autonomous interactions
- [ ] Financial transaction processing
- [ ] System administration tasks
- [ ] External communications
- [ ] Compliance-related decisions

---

## 📈 Section 7: Readiness Scoring

### 7.1 Score Summary

| Section | Max Score | Your Score | Percentage |
|---------|-----------|------------|------------|
| 1. Foundation Prerequisites | 12 | ___/12 | ___% |
| 2. Technical Readiness | 125 | ___/125 | ___% |
| 3. Multi-Agent (if applicable) | 25 | ___/25 | ___% |
| 4. Organizational Readiness | 18 | ___/18 | ___% |
| 5. Governance Enhancements | 21 | ___/21 | ___% |
| **Total (excl. Multi-Agent)** | **176** | **___/176** | **___%** |
| **Total (incl. Multi-Agent)** | **201** | **___/201** | **___%** |

### 7.2 Readiness Level Interpretation

| Score | Readiness Level | Recommendation |
|-------|-----------------|----------------|
| **80%+** (141+/176) | **Ready** | Proceed with agentic AI pilot |
| **60-79%** (106-140/176) | **Preparation Needed** | Address key gaps, then pilot |
| **40-59%** (71-105/176) | **Significant Gaps** | Focus on foundational improvements |
| **<40%** (<71/176) | **Not Ready** | Continue building GenAI maturity |

**Your Readiness Level:** _________________________

---

## 📋 Section 8: 90-Day Agentic AI Readiness Roadmap

### Phase 1: Foundation (Days 1-30)

| Week | Focus | Activities | Deliverables |
|------|-------|------------|--------------|
| 1-2 | Assessment & Planning | Complete this assessment, identify gaps | Gap analysis, roadmap |
| 3-4 | Governance Updates | Update policies for autonomous actions | Policy addendum |

**Key Actions:**
- [ ] Complete full readiness assessment
- [ ] Present findings to AI Ethics Board
- [ ] Update governance framework for agents
- [ ] Identify pilot use case
- [ ] Secure executive sponsorship

### Phase 2: Technical Preparation (Days 31-60)

| Week | Focus | Activities | Deliverables |
|------|-------|------------|--------------|
| 5-6 | Infrastructure | Set up agent development environment | Dev environment |
| 7-8 | Tools & Integration | Implement required APIs, guardrails | Integration layer |

**Key Actions:**
- [ ] Select and configure agent framework
- [ ] Implement monitoring and logging
- [ ] Set up guardrails and safety mechanisms
- [ ] Create development and testing environments
- [ ] Document API integrations

### Phase 3: Pilot & Learn (Days 61-90)

| Week | Focus | Activities | Deliverables |
|------|-------|------------|--------------|
| 9-10 | Build & Test | Develop pilot agent, extensive testing | Working pilot |
| 11-12 | Deploy & Monitor | Limited deployment, close monitoring | Pilot learnings |

**Key Actions:**
- [ ] Build pilot agent with full guardrails
- [ ] Conduct thorough testing (including adversarial)
- [ ] Deploy to limited user group
- [ ] Monitor closely and gather feedback
- [ ] Document lessons learned
- [ ] Plan next phase based on results

### 90-Day Success Criteria

| Criterion | Target | Actual | Met? |
|-----------|--------|--------|------|
| Readiness score improvement | +20% | ___% | ⬜ |
| Governance updates complete | 100% | ___% | ⬜ |
| Technical infrastructure ready | Yes | _______ | ⬜ |
| Pilot agent deployed | Yes | _______ | ⬜ |
| Zero P1/P2 incidents | 0 | _______ | ⬜ |
| User satisfaction (pilot) | >4/5 | ___/5 | ⬜ |

---

## ✍️ Sign-Off

| Role | Name | Signature | Date |
|------|------|-----------|------|
| AI Program Lead | _____________________ | _____________________ | ___/___/______ |
| Technical Lead | _____________________ | _____________________ | ___/___/______ |
| AI Ethics Board Rep | _____________________ | _____________________ | ___/___/______ |
| Executive Sponsor | _____________________ | _____________________ | ___/___/______ |

---

## 📞 Support & Resources

**GenAI Maturity Portal:** https://genaimaturity.net
**Assessment Tools:** https://genaimaturity.net/assessment
**Implementation Resources:** https://genaimaturity.net/implementation

---

_This assessment should be completed before pursuing agentic AI initiatives. Organizations at maturity level 4+ are best positioned for agentic deployments._

**Document Version:** 1.0
**Last Updated:** ___________________________
**Next Assessment Date:** ___________________________