Hi,

I need urgent help recovering data on my Postgres database.

Service: Postgres (host: zephyr.proxy.rlwy.net:22870, database: railway)

Project: shu-abalak

What happened:

While running prisma migrate diff locally against our production

database, we mistakenly passed our REAL production DATABASE_URL as the

--shadow-database-url parameter. Prisma's shadow-database mechanism

treats whatever URL is passed there as disposable — it drops and rebuilds

the schema from migration history to compute a diff. We believe this

caused Prisma to DROP and recreate every table in our production database.

Current state (confirmed via read-only queries just now):

• All 37 tables still exist with the correct schema structure (all 44

  migrations show as applied)

• But every single table is completely empty — 0 rows in users,

  businesses, orders, areas, everything

Before this happened, this database had real data (confirmed by our own

checks in recent days): approximately 31 users, 22 businesses, 28 areas,

and order history including at least 13 cancelled orders, plus normal

day-to-day production traffic.

What I need:

Do you have Point-in-Time Recovery (PITR) or WAL-based recovery available

for this database, even if it's not visible under the "Backups" tab in

the dashboard? I need to restore to a point in time BEFORE today

17.07.2026 17:00 PM, ideally as close as possible to

just before this incident.

This is time-sensitive — please let me know as soon as possible if

recovery is possible, and what I need to do on my end.

Thank you.