Skip to content
PERINGER Data Solutions, back to home

Back

Solution

Connectors between your systems

Between two systems that don’t talk to each other there’s almost always somebody retyping. No-code tools manage for a while, then fail silently the day a source changes format.

The end of double entry

In practice

  • One connector per exchange, with its input format declared and validated
  • Automatic retry on a transient error, and a clean stop on a data error
  • Idempotence: a replayed message does not create a duplicate
  • A readable exchange log, with the payload kept for the agreed period
  • An alert routed to a named person, with what they are expected to do
  • Transformations written as versioned, tested code rather than in a drag-and-drop interface

Systems involved

  • Swiss accounting packages and ERPs
  • CRM and sales tools
  • Online shops and marketplaces
  • Files dropped on SFTP, shared drop folders and mailboxes
  • APIs of the services you already use, where they exist

Service lineApplications and integration →

How it runs

  1. Mapping

    What is retyped by hand at present, between which systems, how often, and by whom.

  2. First link

    The most expensive exchange comes first, read-only at the start, to check what actually arrives.

  3. Writing back

    The connector writes into the target system, with idempotence and retry in place before the first real volume.

  4. Watch

    Alerts routed, log readable, and a one-page procedure for each foreseeable failure.

Test the fit: Connectors between your systems

Describe the context, constraints and decision you need to make. The first conversation qualifies scope, boundaries and the next useful step.

Describe the situation