You thought it was a brain. Let us find out together. A 2.5-hour journey from hoarding to intelligence.
Governance first: workshop assessments stay in the room. No attendee data is uploaded to cloud AI services during the session.
We had 47 tools, Slack channels nobody could keep track of, and a "data lake" the size of an ocean. But when our CEO asked, "What did we decide about the budget in March?" ... three people gave three different answers.
Composite scenario based on patterns observed across workshops.
Assess → Score → Design → Validate. Four phases. One closed loop.
Many teams are a mix of 2s and 3s. That is the gap we will close.
| Level | Awareness | Usage | Integration | Teaching | Score |
|---|---|---|---|---|---|
| 1 · Novice | None | Never | None | No | 0–20 |
| 2 · Beginner | Aware | Not using | None | No | 21–40 |
| 3 · Explorer | Active | Using | Inconsistent | No | 41–60 |
| 4 · Practitioner | Deep | Daily | Integrated | Mentoring | 61–80 |
| 5 · Strategist | Expert | Leading | Systemic | Teaching | 81–100 |
These are diagnostic questions about your current tooling. No data is processed during the workshop. Score each 0–10.
Can an AI query your CRM without a human exporting a CSV first?
Can an AI read meeting recordings and extract action items?
Can an AI search across Slack, email, and docs in one query? (Requires unified search infrastructure. Most organizations need 6–12 months to reach this.)
Can an AI connect a customer complaint to the code commit that caused it?
Can an AI answer "What did we decide about X last quarter?" without asking a human?
Can an AI propose next quarter's priorities based on last quarter's outcomes?
Ticket arrives → AI classifies + suggests response.
Agent reviews → Accepts, edits, or rejects.
CSAT measured → Did the customer agree?
Feedback refines the prompt → Next ticket gets a better suggestion.
Expected result: Response time drops. Agent satisfaction rises. Customer satisfaction rises. Feedback refines future suggestions. No claim of autonomous learning without defined eval metrics and rollback.
Technical note: This describes a retrieval-augmented generation (RAG) loop or rules-based refinement, not model retraining.
The scorecard and canvas are workshop assessments only. Production deployment requires a separate security and compliance review.
The final 20 minutes: Demo Day. Each team presents their closed-loop design. Others score Clarity (1-5) and Feasibility (1-5). Highest total gets first sprint priority.