The first general human behavior API.

Build customer prediction into your product. Send an audience and a decision. Get back what they will choose, what they will pay, and why.

Meta Ads Manager
Predict top ad
Audience pick

2.1× predicted CTR

CREATIVE01 / 06
01 / CREATIVE

Choose the ad that will sell.

Built for ad platforms and generative creative tools like Meta Ads Manager or AdCreative.

Meta Ads Manager
Predict top ad
Audience pick

2.1× predicted CTR

02 / RECOMMENDATIONS

Rank what someone should see next.

Made for streaming and content apps. Think Netflix or Spotify home rails.

Netflix
For tonightRank for me
Best match

2.5× more likely to finish

03 / PRICING

Price the product before it launches.

For commerce back ends like Shopify or WooCommerce admin.

Shopify Admin
Price
$79Predicted
Set the price for me

$79 holds 92% of buyers

04 / PRODUCT UX

Pick the flow customers understand.

At home in design and prototyping tools. Figma, v0, or wherever you vibe-code.

Figma
FLOW A
FLOW BCustomers pick B
Test with customers

85% finish checkout in Flow B

05 / MESSAGING

Send the email they will answer.

Built for outbound and campaign tools like Apollo, HubSpot, or Mailchimp.

Apollo
To2,148 contacts
Subject
EditEdit
Predict the outcome

Two edits lift predicted CTR +38%

06 / CONCEPTS

Choose what deserves to be made.

For research and planning boards. Notion, Airtable, Miro.

Notion
Test all three
51%Make this
31%
18%

51% would buy

07 / RETENTION

Choose the offer that keeps them.

Made for lifecycle and support platforms like Intercom or Braze.

Intercom
Cancel flowPredict the save
On cancel
20% off46%
Pause a month78%Best save
Downgrade61%

The pause beats the discount

08 / MERCHANDISING

Feature the trip people will book.

For CMS and merchandising tools like Contentful or Builder.

Contentful
Homepage heroFill it for me
Empty slot
Candidates
42% would book

42% booking intent

09 / PERSONALIZATION

Show each visitor the right product.

Lives inside storefronts and landing pages. Dynamic Yield, Mutiny.

Shopify Storefront
New visitor · 25–34 · urbanRehearsals API
For you
3.1×

3.1× more likely to buy the lead

10 / EXPERIMENTS

Call the A/B test before it runs.

Built for experimentation stacks like Optimizely, LaunchDarkly, or Statsig.

Optimizely
EXP-042Simulate the test
A31%
B43%Ship B

B wins 94% of simulated runs

API ACCESS

What would you build if you could ask the customer first?

Bring us the product, workflow, or repeated decision you want to put customer prediction inside.