Privacy Policy

1. Information We Process

When using Subtext, you may submit snippets of text messages, chat threads, or workplace communications for linguistic analysis. This text is:

• Processed Ephemerally: Passed in-memory via secure HTTPS directly to the TypeSafe AI Jev System One inference engine to generate gauge scores.
• Never Saved: Not indexed, logged, or recorded in any persistent database or data warehouse.
• Never Used for AI Training: Neither Subtext nor TypeSafe AI trains or fine-tunes models on user inputs.

2. Client-Side Receipt Rendering

When you generate, copy, or download a Subtext Receipt (PNG or vector SVG), all visual rasterization and canvas rendering occurs locally within your web browser:

• Your downloaded images are generated on-device via HTML5 Canvas and vector SVG APIs.
• No image files or screenshots of your receipts are uploaded to or stored on our servers.
• Sharing a receipt to Instagram Stories, TikTok, X, or iMessage is entirely under your control.

3. Demo Presets & Local Caching

The canned demo prompts available under "Try" (such as Per My Last Email or Sunday Breadcrumb) are pre-computed and cached locally in the application bundle. Interacting with demo presets executes with zero network requests and zero transmission of data.

4. Data Security in Transit

All communications between your browser, our servers, and the TypeSafe AI Jev System One API are encrypted using modern Transport Layer Security (TLS 1.3 / HTTPS). We enforce strict HTTP Strict Transport Security (HSTS) headers across all domains.

5. Cookies & Analytics

Subtext does not use tracking cookies, advertising trackers, or cross-site fingerprinting. We do not sell, rent, or monetize user data under any circumstance.