## 🛠️ Core Frameworks & Methodologies

### 1. MEDDPICC (Enterprise Deal Qualification)
**Metrics · Economic Buyer · Decision Criteria · Decision Process · Paper Process · Identify Pain · Champion · Competition**

Apply rigorously:
- Map AI value to **Metrics** the EB cares about
- Tie POC design to **Decision Criteria**
- Use **Identify Pain** to kill "innovation theater" pilots
- Arm Champions with **internal sell kits**

### 2. Technical Discovery Stack

**Layer 1 — Business Outcome**
- What decision or workflow changes?
- What's the cost of the status quo?

**Layer 2 — Data & Integration Reality**
- Sources, volume, latency, quality, labeling, PII/PHI, access controls
- ERP/CRM/data warehouse/API constraints

**Layer 3 — ML/AI Fit**
- Problem type: classification, extraction, generation, forecasting, retrieval, agentic workflow
- Baseline performance, acceptable error cost, explainability needs

**Layer 4 — Operations & Governance**
- MLOps maturity, monitoring, retraining, rollback, audit trails
- Model risk management (especially FS/Healthcare)

**Layer 5 — People & Change**
- End-user adoption, HITL design, training, union/work council concerns

### 3. POC Design Canvas

| Element | Definition |
|---------|------------|
| **Hypothesis** | If we deploy X, metric Y improves by Z |
| **Scope** | In/out of scope datasets, users, integrations |
| **Success Metrics** | Primary (1-2) + guardrail metrics |
| **Duration** | Timeboxed with phase gates |
| **Data Contract** | What customer provides, by when, in what format |
| **Exit Criteria** | Go/no-go to production + commercial path |
| **RACI** | Customer vs. vendor responsibilities |

### 4. AI Solution Reference Architectures

Fluent in articulating patterns:
- **RAG** — retrieval pipelines, chunking, grounding, citation strategies
- **Fine-tuning vs. prompt engineering vs. distillation** — when each is worth the cost
- **Agentic workflows** — tool use, orchestration, guardrails, cost/latency tradeoffs
- **Real-time inference** — streaming, edge, batch, SLA tiers
- **Feature stores & model registries** — enterprise MLOps maturity levels
- **Multi-tenant SaaS vs. VPC vs. on-prem** — security narrative for each

### 5. ROI & Business Case Modeling

Build defensible models with:
- **Value drivers:** labor savings, error reduction, throughput, conversion, fraud reduction, churn
- **Cost components:** inference, storage, labeling, integration, change management, support
- **Sensitivity analysis:** show best/base/worst cases
- **Payback period** and **NPV** framing for Economic Buyers

### 6. Competitive Intelligence Discipline

Structure battlecards around:
- Architectural differentiation (not feature checklists)
- **Landmines** — questions that expose competitor weaknesses
- **Trap doors** — areas not to compete head-on
- **Proof assets** — benchmarks, customer stories, third-party validations

### 7. Demo Excellence

**The 3-Act Demo:**
1. **Pain** — mirror their workflow in 60 seconds
2. **Proof** — show the "aha" with their vocabulary and data shape (synthetic if needed)
3. **Path** — POC → production → expand, with timeline

**Demo rules:** No live gambling. Pre-load fallbacks. Narrate failure handling. End with next step.

### 8. Industry Playbooks (High-Level)

| Industry | Top AI SE Motions | Key Blockers |
|----------|-------------------|--------------|
| **Financial Services** | Fraud, AML, doc processing, advisor copilots | Model risk, explainability, audit |
| **Healthcare** | Clinical doc, coding, prior auth, imaging | HIPAA, bias, liability |
| **Retail/E-com** | Search, recommendations, support agents | Latency, seasonality, margin |
| **Manufacturing** | Predictive maintenance, quality, supply chain | OT/IT convergence, edge |
| **Public Sector** | Citizen services, document intelligence | Procurement, transparency |

### 9. Objection Handling Matrix

| Objection | SE Response Strategy |
|-----------|------------------------|
| "AI isn't accurate enough" | Error taxonomy, HITL, guardrails, baseline vs. target |
| "We can build in-house" | TCO, time-to-value, talent, maintenance burden |
| "Security won't approve" | Architecture options, data residency, zero-retention, audit |
| "No budget" | Phased ROI, start with cost center pain, land-expand |
| "We tried AI before" | Diagnose failure mode: data, scope, change, vendor |

### 10. SE Org Leadership Levers

- Hiring profile: T-shaped — broad architecture + deep AI fluency + executive presence
- Weekly rituals: deal reviews, demo dojo, competitive office hours
- Asset hygiene: versioned battlecards, recorded golden demos, solution templates
- KPIs: win rate, POC conversion, cycle time, SE-attached ARR, customer satisfaction