Workflow automation between business systems

Two systems joined by a drawn line

Connecting two systems used to mean a piece of middleware somebody had to write, host and remember. In STRAX it is a drawn line: a change in one system triggers a step in another, and the relay runs on its own from that moment. There is no code, no server, and nothing to hand over when the developer leaves.

Triggered by a data change, a schedule, an HTTP call or a click.
Runs within seconds and leaves a record of every run.
Changed by dragging a step, with no code to redeploy.

Live Two systems, one drawn line No code · no server · nothing to hand over

A change in one system triggers a step in another, along a line that runs from the moment it is drawn.

Decisions drawn as branches

Real work has conditions: in stock or not, under the limit or over it, first attempt or third. Those decisions are drawn as branches and rules on the diagram, so the process approves what it can by itself and brings a person in only where judgement is needed. When a system is slow or down, a step retries with back-off instead of failing the whole run.

Branches, checks and approvals by rule, all visible on the diagram.
Retries with back-off and error branches, so a slow system does not become an outage.
People are asked only where a decision needs one.

Live Branches and rules on the diagram Auto-approve · retry · ask only for judgement

Branches, checks and approvals by rule, visible on the diagram, with people asked only where judgement is needed.

A nine-step process, watched live

When a process spans five systems (the CRM, billing, provisioning, the support desk and the field app) it is still one diagram, deployed with one click, with every run streaming step by step. You see where it is, what it did and what it is waiting for, and a version moves from staging to production the same way a report does. Runs are shared across a pool of workers, so a busy process scales out by adding workers, while runs for the same customer can be kept one at a time and in order.

One diagram for a nine-step, five-system process.
Versioned deployments to a pool of workers that scales out, promoted from staging to production with one click.
Every run visible live and every failure explained.
A limit on how many runs of a workflow happen at once, fair sharing so one workflow's backlog never holds up the others, and temporary workers an autoscaler can add when the queue grows.

Live Nine steps, five systems, one diagram Streamed live · promoted with one click

A nine-step, five-system process on one diagram, every run streamed live and versions promoted with one click.

Your own code, reused safely

Some steps need code your developers already wrote, such as a pricing rule, or a well-known open-source package for dates, documents or file formats. Your team uploads its own .NET library to STRAX once, or imports a package from NuGet, and the Code steps of every workflow that opts in can use it. Before a package is stored, an administrator sees a preview of it and of everything it depends on, with each licence and a warning for anything deprecated, unlisted or known to be vulnerable.

Only administrators add or change libraries, and every import is written to the audit trail.
Deploying a workflow pins the exact library version it was checked against, until the workflow is deployed again.
Workers fetch libraries from your own STRAX installation and never reach out to a package feed themselves.
Every version is kept, so an earlier one can be made current again, and your own private NuGet feeds can sit beside nuget.org.
The assistant reads each library's public interface, so the Code steps it drafts use your libraries correctly.

Keep exploring