Do I need to be technical to run this loop?

No. The loop is designed for business leaders who can explain the problem clearly and review the answer critically. A technical partner may help with data or systems, but the business owner should make the final decision.

Can I use this with ChatGPT, Claude, or another AI assistant?

Yes. Most sessions work in ChatGPT, Claude, Gemini, Copilot, or an enterprise assistant. Use Codex or Claude Code when the session involves an editable artifact such as a document, spreadsheet, site, repository, or prompt library file. For confidential customer, employee, supplier, pricing, or employer data, use the assistant your organization allows for that material.

When is the loop finished?

It is finished when the decision, next action, owner, evidence gap, and stop condition are explicit. If those are still unclear, rerun the evidence step before adding more people or tools.

Data Readiness Triage Loop | AI Work Session

Data Readiness Triage Loop

By Juan Beltrán — personal website on AI and digital growth for complex B2B industries.

Is the data truly blocking the AI idea, or can we redesign around it? Use this when a team says the data is not ready and leadership needs to know whether that is a blocker, constraint, repair task, or excuse. Data Readiness Triage Loop Task: Is the data truly blocking the AI idea, or can we redesign around it? Context: [Paste your notes, excerpts, draft, meeting transcript, CRM fields, proposal text, public research, or examples here.] Context I should provide: - Use case - Required decisions - Data sources - Data quality concerns - Access constraints - Sample records Useful setup: Paste the AI use case, decision the system must support, available data, gaps, quality concerns, access issues, and examples. Why this matters: Use this when a team says the data is not ready and leadership needs to know whether that is a blocker, constraint, repair task, or excuse. Business problem: AI initiatives stall under the vague claim that the data is not ready, even when only specific data gaps matter. Instructions: Act as a data readiness triage lead. For the AI use case below, identify the minimum data required, inspect the stated quality issues, and classify each gap as blocker, constraint, repair task, or irrelevant. Recommend proceed, redesign, repair, or stop. Workflow: 1. Name the decision: State what the AI system must decide, recommend, generate, or route. 2. List minimum data: Identify only the data needed for that decision, not every possible dataset. 3. Sample reality: Inspect examples for missingness, freshness, consistency, bias, and access. 4. Classify the gap: Mark each issue as blocker, constraint, repair task, or irrelevant. 5. Choose the path: Proceed, redesign, repair, or stop. Quality bar: - Use only the context in this chat. - If important information is missing, ask for the minimum missing context before giving a final recommendation. - Separate facts from assumptions. - Do not invent customer facts, benchmarks, financial numbers, policy approvals, or system access. - Keep the answer useful for AI Product Owner. Output: A proceed, redesign, repair, or stop recommendation tied to the minimum data actually needed. - BLUF recommendation or draft. - Evidence from my context. - Assumptions and missing information. - Risks, objections, or failure modes. - Recommended next action, owner, and stop condition. Evidence checklist: - Decision definition - Minimum data list - Sample records - Quality issue - Impact on output - Owner Stopping condition: Stop when the team can say exactly which data issue blocks which decision.

Key takeaways

Is the data truly blocking the AI idea, or can we redesign around it?
A proceed, redesign, repair, or stop recommendation tied to the minimum data actually needed.
Stop when the team can say exactly which data issue blocks which decision.
Decision definition
Minimum data list

Canonical URL: https://juanbeltran.ch/operating-loops/data-readiness-triage-loop