The Scenario
You are a data engineer and the product team just handed you a Google Sheet with 200 new product records — name, SKU, price, and category for each row — that need to go into the products table of your Prisma Postgres database. The deadline is tomorrow morning so the catalog is live for a product launch.
You could write a migration script. You could set up a local Node process that reads the sheet via API and fires inserts. You could spend the afternoon on either of those things.
The bad version:
- Export the sheet to CSV
- Write a migration script or Node process to read the CSV and fire INSERTs
- Deal with type mismatches on the price column (it came out of the sheet as a string, the table expects a numeric)
- Run the script, get 12 failures on rows with special characters in the product name, fix the escaping, re-run
- Verify the count in the database, notice rows 45 and 46 are missing, trace back the failure
A two-hour job at minimum. And it has to be redone if the sheet changes before launch.
The Easy Way: One Prompt in SheetXAI
SheetXAI is an AI agent that lives inside your Google Sheet. It reads the spreadsheet data directly, and through its Prisma integration it can execute parameterized INSERT and UPDATE statements against your database without a migration script.
With product data in columns A through D, try:
Read each row from columns A–D (name, sku, price, category) and run bulk INSERTs into the products table of my Prisma database — use parameterized queries for each row
What You Get
- One INSERT executed per data row using parameterized queries — no SQL injection risk
- Rows with type mismatches flagged with a specific error note rather than silently skipping
- A count of successful inserts written at the top of the sheet when the batch completes
- Rows that failed are highlighted with the specific Prisma error so you can fix and re-run
What If the Data Is Not Quite Ready
Some rows already exist in the database — I want upserts, not duplicate inserts
Read each row from columns A–D (name, sku, price, category) and run INSERT INTO products ... ON CONFLICT (sku) DO UPDATE SET name=EXCLUDED.name, price=EXCLUDED.price, category=EXCLUDED.category against my Prisma database
I need to UPDATE existing records by ID rather than insert new ones
Use the user IDs in column A and new plan values in column B to run 'UPDATE users SET plan=? WHERE id=?' for each row against my Prisma Postgres database — write "updated" or the error message to column C for each row
The price column has currency symbols that need stripping before insert
Read columns A–D (name, sku, price, category) — strip any currency symbols from column C before inserting, and run bulk parameterized INSERTs into the products table of my Prisma database
Read all rows, upsert to the database, verify the count matches, and flag discrepancies in one shot
Read all rows from columns A–D (name, sku, price, category), run INSERT INTO products ... ON CONFLICT (sku) DO UPDATE for each row against my Prisma database, then query SELECT COUNT(*) FROM products to verify the total, and in a summary row at the top of the sheet write the inserted count, updated count, and failed count
The pattern: drive the entire write operation from the sheet in one prompt, including the verification step, so you know the database is in the right state before the launch deadline.
Try It
Get the 7-day free trial of SheetXAI and open any sheet where you are staging records that need to land in a database table, then ask it to run the bulk inserts against your Prisma Postgres database. See also the spoke on running SELECT queries into a sheet or the hub overview.