Action

Router

Branch on extracted-data rules

Router branches a job based on the actual values pulled from a document, sending each one down the path that matches rules like amount over a threshold or state equals CA.

How it works

Router is where a flow makes a turn based on what the paperwork actually says. After the values are extracted, Router checks them against the rules you set (loan amount over 50,000, applicant state is California, revenue below a floor) and sends the job down the matching path.

This is the difference between Router and Classify. Classify branches on fixed rules about the job as a whole; Router branches on the real numbers and fields extracted from the document, so your paths respond to the content of each submission, not just a category.

You use it whenever different values deserve different handling: big deals go to a senior reviewer, out-of-state applications take a different route, small amounts fast-track. Each job leaves Router on exactly one path, so the steps after it can assume they are seeing the right kind of case.

Common use cases

Real situations where this node earns its place in a flow.

Send big requests to a manager

Branch on an extracted figure so large loan amounts go to senior review and small ones auto-proceed.

Route by state or region

Send each job down a path based on where the applicant is located.

One workflow, many endings

Use extracted values to decide which delivery or document steps run for a given job.

Router in a workflow

A simple example of where this node sits in a flow.

Extract PDFs → structured data Router Branch on extracted-data rules Decline Reject applicant + decline email

What you set up

  • Pick the extracted fields your rules will read (amount, state, date, count).
  • Write each branch as a condition and point it at the path that should handle it.
  • Set the default path for jobs that match none of your conditions.

Common questions

How is Router different from Classify?

Classify branches on fixed rules you write about the job; Router branches on the actual values extracted from the document, so you can route on things like a dollar amount or a state code.

What if a job matches no rule?

It follows the default path you set, so nothing gets stuck. It is worth pointing that default at a review step so you can see cases your rules did not anticipate.