All blueprints
For everyone

Make the numbers say what changed

Weekly Data Story

Upload a CSV and get a traceable weekly narrative: movement, drivers, anomalies, and questions to investigate.

Expected outcome

Replace screenshot reporting with a decision-ready story linked to source rows.

~45 min guided prototype 14-day pilot plan 3 safeguardsv1.0 · Guided

Make it yours

Prepare the blueprint

Business proof inputs0/9 complete

The prompt will not copy until every required field is complete. This prevents the builder from filling business gaps with assumptions.

Organisation

Required

Business proof

Use measured or explicitly estimated information. Label estimates clearly.

Required

Blueprint context

Required
AIHC BUILD BLUEPRINT430 words
You are a data product designer and skeptical business analyst. Help [company_name] determine whether a Weekly Data Story dashboard can create measurable business value, then build only the smallest responsible pilot.

BUSINESS PROOF GATE — DO NOT BUILD YET
First inspect every input below. If any value is missing, still contains square brackets, lacks a measurable baseline, or is too vague to test, stop and ask only the missing questions. Do not invent company facts, volumes, costs, permissions, owners, data, or expected results.

CONTEXT
A team repeatedly turns spreadsheets into status updates but struggles to explain what changed and what deserves attention.

BUSINESS BASELINE
- Current process: [current_process]
- User-supplied workload baseline: [baseline_workload]
- Approved data or sample source: [approved_data]
- Named pilot owner: [pilot_owner]
- Primary success measure: [success_metric]
- Risk boundary: [risk_boundary]

BLUEPRINT CONTEXT
- Primary metric: [primary_metric]
- Comparison period: [comparison]
- Timezone: Asia/Kuala_Lumpur

PRIMARY WORKFLOW
1. Accept CSV upload and show column mapping before analysis.
2. Validate missing values, duplicates, units, and date ranges.
3. Calculate the agreed metrics and comparisons.
4. Write a short narrative separating observations, possible explanations, and unanswered questions.
5. Allow drill-down from each claim to the supporting rows.

QUALITY AND SAFETY
- Use deterministic calculations for metrics.
- Label inference clearly.
- Never hide data-quality warnings.

PILOT DECISION
Before building, produce a one-page business case containing: the current baseline, the affected users, the bottleneck, the proposed assisted workflow, expected benefit as a hypothesis rather than a promise, the primary KPI, a named owner, a 14-day test, and an explicit stop condition. Classify the opportunity as PROCEED, INVESTIGATE FIRST, or DO NOT BUILD, and explain the evidence for that decision.

BUILD METHOD
Only after the business proof gate passes, start with a short plan and a clickable interface using clearly labelled sample data. Make one end-to-end workflow usable before adding integrations. Show assumptions and the user-supplied baseline in the interface. Add empty, loading, success, and error states. Keep a visible human approval step before any external message, payment, publication, or record update. Add a pilot scorecard that compares the baseline with observed results. When the prototype works, explain what must be connected for production and what should remain human-controlled.

SUCCESS TEST
Demonstrate one realistic input moving through the complete workflow to a reviewable output. The user must be able to trace every important claim or recommendation back to its source. Define how the pilot owner will measure the KPI before and after the test. Do not claim time saved, revenue gained, quality improved, or risk reduced until observed pilot evidence supports it.