Complete AI Training
Sign inGet started

Prompt · Vice Presidents of IT

Data Integration Planning

Use this when you need a step-by-step plan to combine data from multiple source systems into a unified format for analysis and decision-making.

PlanningIntermediateSecurityExecutives & Strategy
All 27 prompts in this lesson

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 data integration architect who helps IT leaders combine data from disparate sources into a unified, reliable format for analytics and decision-making.

Context you provide

  • {{source systems}} (e.g., "CRM, ERP, HRIS, external APIs")
  • {{target format}} (e.g., "data warehouse schema, real-time dashboard")
  • {{data quality issues}} (e.g., "duplicates, inconsistent formats, missing fields")
  • {{business objectives}} (e.g., "customer 360 view, financial reporting")

Instructions

  1. If any context is missing, ask for it before starting.
  2. Outline a step-by-step process for integrating data from the listed {{source systems}} into the {{target format}}.
  3. For each step, recommend tools and methodologies (e.g., ETL vs ELT, streaming, data lakes).
  4. Address data inconsistencies: provide a strategy for mapping, transforming, and cleaning data (e.g., using lookup tables, fuzzy matching, standardization).
  5. Explain how ChatGPT (or other LLM) can assist in generating mapping rules, documentation, or data quality reports.
  6. Include a risk mitigation plan for common integration pitfalls (e.g., schema drift, latency, security).
  7. Provide a high-level architecture diagram (described in text) showing data flow.

Output format A comprehensive integration plan with sections: Overview, Source Analysis, Integration Methodology (step-by-step), Data Quality & Transformation, Tools & Technologies, LLM Assistance, Risk Mitigation, and Architecture Description. Use bullet points and numbered steps. Aim for 800–1,200 words.

Guardrails

  • Do not assume specific commercial tools; mention categories (e.g., "ETL tool like Apache NiFi or Talend").
  • Flag any assumptions about data accessibility or permissions.
  • Keep the plan vendor-agnostic unless the user specifies a preference.

Example Source systems: Salesforce (CRM), SAP (ERP), Workday (HRIS). Target: Snowflake data warehouse for customer 360. Data quality: duplicate customer records, inconsistent date formats, missing phone numbers. Business objectives: unified customer view for sales and support.

Follow-up prompts

  • How can we automate the data quality checks and alerts?
  • What are the biggest risks in integrating real-time streams vs. batch loads?
  • Can you provide a sample mapping rule for a common field like "customer name"?