Question

A developer needs a realistic dataset to demo a customer-analytics dashboard. They have a production `customers` table schema (name, email, phone, signup_date, lifetime_value) but company policy forbids copying real production records into non-production environments. They open the schema file and ask Copilot Chat to generate 200 sample rows. Which approach best satisfies both the demo need and the policy?

Accepted Answer

Prompt Copilot to generate 200 rows of entirely synthetic data that conform to the schema's column types and plausible value ranges, with no real production values included.

Answer

Paste 50 real production rows into the chat as few-shot examples so Copilot can produce statistically identical data, then delete the pasted rows afterward.

Answer

Ask Copilot to connect directly to the production database and sample rows, since Copilot only reads and does not persist the data.

Answer

Have Copilot generate a script that copies the production table and lightly shuffles the email column so the records are technically different.