Security & Client-Side Architecture
How MiroMint DocSuite processes documents: by executing logic locally in your browser sandbox without remote server ingestion.
| Security Dimension | Standard Cloud PDF Services | MiroMint DocSuite (Client-Side) |
|---|---|---|
| File Transmission | Uploaded over internet to multi-tenant cloud servers | 0 bytes transmitted over network |
| Server Storage Risk | Stored on AWS S3 / Google Cloud temporary buckets | Zero server storage. No cloud database. |
| Execution Environment | Shared remote Linux containers / EC2 instances | Client browser sandbox (V8 / JavaScriptCore) |
| Breach Surface Area | Massive (centralized repository of millions of files) | Zero Server Ingestion (no central repository to breach) |
| Processing Runtime | Requires continuous internet connection to cloud server | Client-side JavaScript & Web Workers in browser sandbox |
1. Client Sandboxing via Local In-Browser Processing
MiroMint executes binary processing libraries directly in your local browser sandbox using client JavaScript engines, background Web Workers, and GPU-accelerated canvas primitives:
- `pdf-lib`: Handles binary PDF document generation, byte manipulation, stream parsing, and metadata stripping completely in local client memory.
- `pdfjs-dist` (Mozilla Web Worker): Compiles PDF page rendering and text extraction in an isolated background web worker thread (`pdf.worker.min.mjs`), ensuring your main UI thread remains responsive with zero network payloads.
- HTML5 Canvas Compositor: Biometric passport photo cropping, dimension calibration, and ID card stitching execute in local GPU-accelerated canvas memory.
2. Healthcare & Financial Privacy Considerations
In healthcare and financial environments, uploading sensitive files to cloud third parties creates data processing risk and requires Business Associate Agreements (BAAs).
MiroMint functions strictly as a client-side web utility: Because MiroMint has no server-side document ingestion endpoint and does not receive, transmit, or store documents on cloud servers, document bytes remain isolated within your local browser runtime.
3. Permanent Pixel-Burn Redaction
Unlike poorly designed PDF editors that merely place a black box overlay above selectable text (leaving the underlying text selectable by copy-paste or screen readers), MiroMint’s Permanent Redactor converts the PDF page into a rasterized bitmap, applies permanent black pixel fills, and flattens the canvas before re-encoding. The original text stream is destroyed permanently.
4. Session Memory Purge & Lifetime Lifecycle
When an operation is complete:
- Generated download URLs are created via temporary
URL.createObjectURL(blob)references that live only in the current DOM session. - Clicking the "Session Memory Purge" button executes
URL.revokeObjectURL()and clears all canvas 2D contexts, freeing memory back to standard browser garbage collection.
5. How to Verify Zero Server Uploads in Your Browser (DevTools Proof)
You do not have to take our word for it. You can independently verify that your documents never leave your computer in under 30 seconds:
- Press F12 (or right-click anywhere and choose Inspect) to open Developer Tools.
- Click the Network tab, check the Preserve log box, and filter by Fetch/XHR.
- Drop any document or image into MiroMint and compress, edit, or crop it.
- Inspect the network log: you will observe that 0 network payloads contain your document bytes or file contents. All processing takes place in memory on your device.