On-prem data environments
Writable, masked, prod-like clones for every developer, pipeline, and demo. Running in your cloud or datacenter, so the data never leaves.
Trusted by world class organizations
Deploy once. Unblock every team.
Deploy inside your network
The server and connector are Docker images. Installs on your Kubernetes cluster and works with all clouds.
Connect your databases
PostgreSQL, MySQL, MongoDB, and DocumentDB, all under one control plane. Masking happens during the copy, and automated PII scanning is included on-prem.
Every team gets prod-like data
Developers debug on real distributions. CI tests every migration on a fresh clone. QA and demos stop sharing one stale environment. AI agents get a database they can break. Each clone is writable and ready in under 2 seconds.
Real data for every team. In your cloud.
Teams adopt Baseshift to build and test on prod-like data. On-prem means that happens inside your network, at full speed.
Test on real data
Seed scripts and stale staging copies approve anything. Production distributions, NULLs, and tenant skew are where bugs live, and every clone carries them. Migrations, bug repros, and demos behave like production because the data does.
A clone for everyone, in seconds
Under 2 seconds, writable, and unmetered on your own infrastructure. Dev, QA, CI, demo environments, and AI agents each get their own database instead of queueing for staging.
Masked, audited, inside your boundary
Masking and PII scanning happen during the copy, on your infrastructure. Commands, logs, and audit metadata flow to the control plane. Table contents never leave your network.
Enterprise cloning,
without the data leaving home.
The data plane is yours. Baseshift runs the control plane, so your team does not have to.
PostgreSQL, MySQL, MongoDB, and DocumentDB, self-hosted in your network.
On-prem clones typically start in under 2 seconds, on your own infrastructure.
Replication-based snapshots for Postgres and MySQL ship each day's changes as a thin block-level layer, integrated live when a clone runs. On-prem exclusive.
Automated detection of sensitive columns, on top of column-level rules and policy templates.
Optionally expose clones on your existing tailnet, so laptops and CI runners reach them without extra VPN plumbing.
Baseshift copies from your sources with a read-only role. Nothing in the deployment can write to production.
Keep the data. Start today.
The free tier is the same on-prem: 150 GB of snapshots, and clones are unmetered in your cloud. Pro from $30/month when you need more.
Get started for freeSee pricing →FAQ
Does my data leave my network?
No. The server, connector, and every clone run inside your network. The control plane receives commands, logs, and audit metadata; table contents never leave.
Which engines are supported on-prem?
PostgreSQL, MySQL, MongoDB, and DocumentDB.
Is on-prem a different product?
No. It is the same Baseshift, deployed in your environment. Everything the SaaS does, plus MySQL, MongoDB, and DocumentDB support, sub-2-second clones, delta snapshots, and unmetered clone counts.
How do I deploy it?
The server and connector are Docker images. Installs on your Kubernetes cluster and works with all clouds.
How fresh are the clones?
Sources are kept in sync by replication. For Postgres and MySQL, on-prem can also build delta snapshots, where each day's changes ship as a thin additional layer on top of the base snapshot.
Can it run air-gapped?
No. The data plane is fully local, but the control plane runs in Baseshift's cloud, so outbound connectivity is required.
How do developers and CI reach the clones?
Any network path you already use. Optionally, Baseshift integrates with your existing Tailscale network and exposes clones on your tailnet, so laptops and CI runners connect without extra VPN setup.
What does on-prem cost?
The free tier matches SaaS: 150 GB of snapshots. On-prem clones are not metered, so run as many as your infrastructure allows. Talk to sales for enterprise terms.