Skip to content

Workflow engine

Highway

Available

Processes that survive a crash.

A workflow engine for processes that run for minutes, days or weeks and must not be left half finished. The work and the record of the work are committed together, so an outage cannot leave an order paid and undelivered, or a case closed with a step skipped. It runs on the database a team already operates, so there is no separate queue, broker or scheduler to licence, run and patch.

Who it is for. Teams running multi-step processes where a partial failure has a cost: payments, onboarding, fulfilment, case handling and anything with a deadline attached.

Status
Available
Part of
Offerings
Product site
highway.solutions
Tags
Durable executionNo extra infrastructureMulti-tenantAudit trail

What it does

  • Nothing is left half done

    The work and its record commit together. A failure rolls back to a state the business recognises, and the process resumes from there.

  • No new infrastructure

    The PostgreSQL database you already run is the only moving part. There is nothing else to licence, operate, monitor or patch.

  • Durable execution

    Workflows survive crashes and restarts, and resume where they stopped without losing any work.

  • Processes written as code

    Each process is written as readable code, so its steps, retries and approvals are reviewed and versioned like any other change.

  • Waits for people and for events

    Approval steps, webhooks, and fan-out and join patterns are built in.

  • Circuit breakers

    Failures are detected and contained automatically so they cannot cascade across services.

  • Tenant isolation

    Each tenant’s data is isolated, and the rule is enforced inside the database, underneath the application.

  • Audit trail

    Every step is recorded permanently, and past state can be read back as it stood at any moment.

Documentation and access details are on the product site, highway.solutions. For a pilot, an integration or procurement questions, contact us.