AI Condition Builder

The AI Condition Builder lets you describe your targeting logic in plain language and automatically converts it into editable Userflow condition blocks — no manual condition setup required.

How it works

Instead of manually selecting attributes and operators to build your targeting conditions, you type a description of who should see your flow or checklist. The AI Condition Builder interprets your description and generates the matching condition blocks, which you can review and edit before saving.

Example: Type "Users who signed up in the last 7 days and haven't completed onboarding" and the AI Condition Builder produces the corresponding attribute and event conditions.

Using the AI Condition Builder

1. Open a flow or checklist in the builder and navigate to the Auto-start section.

2. Click AI Condition Builder (or the AI icon next to the conditions area).

3. Type a description of your target audience in plain language.

4. Review the generated condition blocks. Each block is fully editable — you can adjust operators, values, or remove conditions individually.

5. Click Apply to save the conditions.

Refining conditions conversationally

You can add follow-up instructions to modify the generated conditions without starting over. The AI Condition Builder appends or overwrites conditions based on your follow-up.

Example: After generating an initial set of conditions, type "Also make sure they are on the Pro plan" — the builder adds a plan attribute condition to the existing set.

Editing generated conditions

All conditions produced by the AI Condition Builder are standard Userflow condition blocks. You can:

• Edit any value directly in the condition block

• Remove individual conditions using the × icon

• Add additional conditions manually using Add condition

• Switch between AND / OR logic as needed

Tips for effective prompts

• Be specific about attributes you track: "Users with plan = starter and days_since_signup < 30" produces more accurate results than "new users."

• Mention the action you want to trigger, not just the audience: "Users who haven't seen the onboarding flow yet" is clearer than "new users."

• If the output isn't right, refine it conversationally rather than starting over.