Skip to main content

Database Change AutomationRequested. Reviewed. Deployed. Audited.

Database changes run through a workflow — request, review, deployment, audit. The 6 levels show how teams get there, and where most plateau.

Database automation maturity

Six levels. From manual to integrated.

Borrowed from the levels of autonomous driving. Each level removes a class of manual work and exposes the next bottleneck.

  • L0

    No automation

    Direct DB clients. Every change typed and run by hand.

  • L1

    Ticketing

    Jira or ServiceNow captures the request, approval, and audit trail. SQL is still executed by a human.

  • L2

    Version controlled

    Schema history and change scripts live in Git. Deployment is still manual.

  • L3

    Streamlined

    Liquibase, Flyway, or sqitch run inside CI/CD. Deployment automated, cross-tool stitching still manual.

  • L4

    Integrated

    Change, ad-hoc access, and audit in one platform. The current ceiling in production.

  • L5

    Fully automated

    End-to-end including self-healing. Aspirational.

Database automation in Bytebase

Change, access, and audit in one workflow.

Schema changes run as pull requests. SQL review, lint, and policy checks run before execution. Ad-hoc access flows through the same approvals. Every action — change, query, export — lands in one audit log. ServiceNow and Jira plug in for CAB approvals. GitHub, GitLab, Bitbucket, and Azure DevOps plug in for execution.

One platform. Every change.

Database automation questions

Common questions.

What does it mean to automate database change?
Automating database change means request, review, deployment, and audit run as one structured workflow — not as ad-hoc tickets, chat messages, and SQL clients. Hand-off points are recorded. Humans gate risk; the system executes.
Where do most teams plateau on the automation curve?
Most teams plateau at Level 3 — schema in Git, migrations applied by Liquibase or Flyway through CI/CD. Change is automated; ad-hoc access, audit, and policy enforcement still live in separate tools. Level 4 consolidates them into one workflow.
Can we keep using ServiceNow or Jira and still automate?
Yes — and most enterprises do. ServiceNow or Jira owns the change request and CAB approval. An automation layer (CI/CD pipeline or change-management platform) owns execution, policy checks, and the audit record. Webhooks or APIs link the two so the ticket reflects what ran.

Explore the standard for database development