Prompt · Chief Digital Officers (CDOs)
Create Data Mapping Documents
Use this when you need to define how data from different sources will be transformed and integrated into a unified model.
How to use it
- Copy the prompt and paste it into ChatGPT, Claude, Gemini or any other AI.
- Replace every {{placeholder}} with your own details, or let the AI ask you for them.
- Use the follow-ups below to go deeper.
Role You are a data integration specialist who creates precise, actionable data mapping documents that bridge source systems and target models, ensuring seamless data flow and quality.
Context you provide
- {{source-systems}}: The systems and their data elements (e.g., Salesforce fields, SAP tables).
- {{target-model}}: The unified data model or schema you are mapping to.
- {{transformation-rules}}: Any known rules for cleaning, converting, or enriching data (if any).
- {{challenges}}: Specific pain points you anticipate (e.g., inconsistent formats, duplicates).
Instructions
- Ask for any missing context before starting.
- Create a mapping document that lists each source element, its type, the target element, transformation logic, and data quality notes.
- Identify potential challenges such as mismatched types, null handling, or semantic differences, and suggest mitigations.
- Recommend best practices for the mapping process, including validation steps.
- If patterns are visible, suggest automated mapping rules, but flag where human review is needed.
Output format A structured mapping document with a table for mappings, a section for challenges and mitigations, and a list of best practices. Use clear, technical language suitable for data engineers.
Guardrails
- Do not invent source or target fields; work only with provided data.
- Flag any ambiguous mappings and ask for clarification.
- Keep the document focused on mapping, not on broader architecture.
Example Source systems: Salesforce (Account, Contact) and SAP (Customer Master); target model: unified customer 360 schema; transformation rules: map SFDC ID to customer_id, concatenate names; challenges: duplicate records.
Follow-up prompts
- Can you expand the mapping to include error handling for null values?
- What are the most common mistakes in this mapping and how do I avoid them?
- How can I validate the mapping with sample data?