About SafeGrd

Why we built it

SafeGrd was founded by an engineer who worked on data security and storage at Google and Amazon, including encryption at rest and key management. In that work, the backups that failed were usually ones nobody had tried to restore. When someone finally needed them, they were missing tables, had been cut short, or were encrypted to a key nobody had kept.

Large companies avoid this by running restore tests on a schedule and keeping backups where production credentials cannot reach them. Most smaller teams have neither, and that matters more now that AI agents and deploy scripts hold production credentials, because a token that can drop a database can often delete its backups as well. SafeGrd gives those teams both.

What it does

SafeGrd backs up PostgreSQL databases, file trees and IMAP mailboxes. Each backup is encrypted on your host and written to a bucket under S3 Object Lock, so it cannot be deleted before its retention ends. On a schedule, SafeGrd restores the newest backup, compares its tables, rows and file digests with what was backed up, and signs a record of the result, which you can show to a customer or an auditor.

Where we are

SafeGrd is based in Bengaluru, India - 560048. The legal entity is named on the contact page.

If SafeGrd stops

Your backups are in your bucket (or ours, with an export to yours at any time), and the CLI restores them without an account. Its source is public at github.com/safegrd/cli, and its licence allows all production use if SafeGrd stops running its service. See the FAQ and Trust and security.

Contact

Write to support@safegrd.dev for support and billing. The changelog lists what has shipped so far.