One node, every path: meet the Switch node

Route each result in CloudFlow to the right action based on a field's value, no more stacking Branch nodes to handle multi-way decisions.

Real cloud operations rarely come down to a single yes-or-no question. A tag can hold five different values. A request can name any of your providers. A finding can carry one of a dozen categories.

Until now, modeling "do something different for each value" in CloudFlow meant chaining Branch nodes — one if/else after another. It worked, but flows grew deep and repetitive, adding a new value meant re-wiring the chain, and understanding the logic at a glance got harder with every branch.

What's new

The Switch node turns that chain into a single decision point. Point it at one field, name a case for each value you care about, and CloudFlow routes every result down its matching path:

Named cases with exact matching - match strings, numbers, or booleans, with strict typing so "1" and 1 never get confused.

Ignore case — one checkbox makes Production, PRODUCTION, and production land on the same path. No more normalizing tag values upstream.

Default path — catch everything that doesn't match a case, or leave it off and unmatched results simply end there. Nothing fails, nothing falls through by accident.

Per-result routing — this is the part we're most excited about. The Switch node evaluates each item in a batch independently, so one run can activate several paths at once, with every item going exactly where it belongs.

Example: After-Hours Instance Sweep

Imagine a flow that runs every weekday at 19:00, lists all EC2 instances, and hands the batch to a single Switch node that reads each instance's Environment tag:

• production → skipped, untouched. Guardrail built in.
• dev → stopped for the night. That's idle spend you stop paying for, automatically.
• Everything else → the default path looks up the resource owner and asks them to tag it, so your untagged inventory shrinks over time instead of accumulating.

One scheduled trigger, one list call, one routing decision. The same sweep with Branch nodes would take a cascade of conditions; with Switch it reads exactly like the policy you'd write on a whiteboard. And because routing is per-result, a single evening run stops the dev boxes, protects production, and nudges owners about untagged instances, all in one execution.

Get started

The Switch node is available now for all CloudFlow users. In the flow editor, select [+] where you want to route results, choose Switch, pick the field to evaluate, and add your cases. Check out the Switch node documentation for matching rules and more examples.

If a field in your cloud can hold more than two values, there's now a node for that.