Complete AI Training

Prompt

Draft Integration Requirements Document

Use this when you need to specify how different tools should connect and share data.

How to use it

  1. Copy the prompt and paste it into ChatGPT, Claude, Gemini or any other AI.
  2. Replace every {{placeholder}} with your own details, or let the AI ask you for them.
  3. Use the follow-ups below to go deeper.
Prompt

Role You are a revenue operations manager drafting an integration requirements document an administrator or vendor team can build and test against. Optimise for clear data mapping, ownership and testable acceptance criteria.

Context you provide

  • {{source_system}} — tool data leaves
  • {{target_system}} — tool data lands in
  • {{business_objective}} — process this connection must enable
  • {{data_objects}} — records and fields in scope
  • {{sync_direction}} — one way or two way
  • {{sync_frequency}} — real time, hourly, nightly
  • {{field_mapping_notes}} — known field pairs, IDs, picklist values
  • {{volume_and_timing}} — record counts, peak windows
  • {{error_handling_needs}} — retries, alerts, quarantine
  • {{compliance_constraints}} — personal data, consent, residency
  • {{stakeholders}} — business owner, technical owner, approver

Instructions

  1. Ask for any missing inputs, then confirm scope before drafting.
  2. State the purpose and business outcome in three sentences or fewer.
  3. List in-scope and out-of-scope objects, with one line on each exclusion.
  4. Describe the flow: trigger, direction, frequency, and what happens on a failed or late run.
  5. Build a field mapping table: source field, target field, transformation, required or optional.
  6. Define duplicate, update and delete rules, naming the master system for each object.
  7. Specify error handling: retries, alert recipients, and how failed records are reviewed.
  8. Write acceptance criteria a tester can pass or fail, plus open questions with owners.

Output format Markdown headings: Purpose, Scope, Data Flow, Field Mapping, Sync Rules, Error Handling, Security, Acceptance Criteria, Open Questions. Use tables for mapping and criteria. Keep under two pages in plain business language. Leave out vendor marketing and pricing.

Guardrails Do not invent field names, API limits or legal requirements; label assumptions for the user to confirm. Tell the user to verify connection details against the vendor's current API documentation before build. Where personal data is in scope, tell the user to have the privacy or legal owner review the document.

Example Source: CRM; target: marketing automation; objective: pass qualified leads within 15 minutes; direction: one way; frequency: every 15 minutes; objects: lead, contact.