Workflows

Workflows are automated decision logic flows, that can power customer risk decisions for account opening, ongoing monitoring, fraud checks, credit decisions, and more. Let's build a very simple workflow together.

Each workflow begins with an Input and runs from left to right. You can drag and drop Logic Blocks into your Workflow, to build all sorts of decision logic - these blocks can include Rulesets, Data Connectors, Transforms, and more

Input and Fields

Each Workflow expects a series of data fields in its Input. These fields can be accessed in your rules, to make decisions and calculations in your Workflows.

For each workflow, you Input will have an object name (e.g. customer or transaction) as well as a list of fields associated with that object (e.g. customer.email_address)

You can customize the fields for your Workflow by clicking on the Input block. Or you can simply pass new fields to the Input block via API.

Outputs

Every time your Workflow runs, it will return:
- Result (Approve, Reject, Investigate)
- Custom Variables (optional) such as score, risk_tier, etc

End Points are red dots that show at the end of each Path of your workflow.

You can click on an End Point to see the Outputs returned when your workflow follows this Paths (see [Paths and Actions](Paths and Actions) for more details). You can also adjust which Custom Variables you actually want to return by flipping the toggles on this screen.

Data

In addition to the data fields passed into the Workflow's Input, you can connect your Workflow to additional data sources. Each time your Workflow runs, it can then call out to these data sources, and pull back additional fields to use in your rules.

Sliderule supports

  • API Connectors - for calling external data vendors, or internal services at your company
  • Database Connectors - for querying SQL-style databases, and pulling back specific columns as fields
  • GRPC Connectors - for fast connectivity in our Enterprise deployments

You can create and manage the credentials used for each data connection on the Credentials page.


Did this page help you?