This section talks about the underlying mechanics of Magic Patterns, shedding light on the technology that powers our service, the architecture, and the measures we take to ensure data security.

Familiarity with this section isn’t necessary for the use of Magic Patterns, but we offer this information for those interested in the technical background and for comprehensive security evaluations.

High-Level How Magic Patterns Works

1

User prompt

You begin by entering a text prompt describing the user interface you envision. This could be as simple as “login page for an app” or as specific as detailed descriptions of elements and their behavior.

2

Retrieve relevant context

Upon receiving your prompt, Magic Patterns retrieves the necessary context from its extensive library of design elements and systems. This process involves matching your description to the most appropriate UI components and layout patterns.

3

Call relevant artificial intelligence model

The AI model is then engaged to interpret the prompt and context, generating a UI design. The model has been trained on a wide array of UI patterns and design principles, enabling it to produce high-quality, functional designs. Enterprise customer data is not used to train or improve any models.

4

Show output

The final step is showing the output. The generated UI is rendered into a visual preview for your review and can be exported to code or to Figma, ready for implementation or further refinement.

Subprocessors

For comprehensive compliance information including SOC 2 reports, security policies, and more, visit our Trust Center.

VendorNature and PurposeSecurity and Supplemental Measures
AnthropicAnthropic is an AI safety and research companyhttps://trust.anthropic.com/
PostHogPostHog provides analytics and product insightshttps://posthog.com/docs/privacy
SentrySentry provides application performance monitoring and error trackinghttps://sentry.io/security/
CloudflareCloudflare provides content delivery and security infrastructurehttps://www.cloudflare.com/trust-hub/
MongoDBMongoDB is a database providerhttps://www.mongodb.com/products/capabilities/security
RenderRender is a platform-as-a-service (PaaS) used to host and run application serviceshttps://render.com/security
VercelVercel is a platform-as-a-service (PaaS) used for frontend hosting and deploymentshttps://vercel.com/security
Google WorkspaceGoogle Workspace is used for business productivity and internal collaborationhttps://workspace.google.com/security/

Security

Read about Magic Patterns’ security measures and how we protect your data.