How Vettamed limits protected-content exposure.
Vettamed splits itself in two on purpose: the web workspace excludes protected question bodies, while the desktop application receives an encrypted session manifest and keeps decrypted content in process memory for the active session. This page explains that boundary and its limits.
Two surfaces, one boundary
Vettamed Web handles accounts, plans, progress, classroom, and billing. Its product queries and responses are designed around safe metadata such as lifecycle state, subject, and difficulty; protected prompts, options, explanations, and answer keys are not part of the web product surface.
Vettamed Desktop is where protected content renders. Each session negotiates a fresh, ephemeral encryption key with the server; the decrypted manifest stays in application memory for the active session and is not intentionally written to disk, browser storage, application logs, crash reports, or analytics. Closing the session clears renderer state and key material.
This is a tested architectural boundary, not a claim that endpoint compromise is impossible. It reduces casual extraction and limits where plaintext can exist while keeping the protected bank out of ordinary browser workflows.
Two modes inside Desktop
Not every study session needs the same posture. Vettamed Desktop runs in one of two protected modes, and the interface tells you honestly which one you are in.
Used for ordinary practice, assignments, and mistake review. Content is encrypted at rest and in transit, then held in memory after authorised decryption; the interface stays light and does not lock your machine down.
Used for full mocks and live mock days. It adds the same protected-delivery controls to a locked kiosk-style interface, sealed submission, and session-level device binding for a more exam-like rehearsal.
If content leaks anyway
Software cannot stop someone from photographing their own screen with a second device. We do not pretend otherwise. What we built instead is attribution: every protected render carries an invisible-to-casual-inspection forensic marker, unique to that session and that question.
If a leaked screenshot surfaces, an administrator can trace it back to the specific session it came from - after re-authenticating and creating an immutable audit record of the lookup. The marker never encodes your name or IP address directly; those are recovered only through the protected server record, under that same audited process.
This is why the question bank stays valuable at scale: not because leaking is impossible, but because it is traceable, which changes the incentive to try.
This is leak deterrence and attribution. It is not camera-proof, and we will never describe it as such.
What Vettamed does not do
- No webcam access. We never request, capture, transmit, or store any image from your camera.
- No microphone access. We never request, capture, transmit, or store any audio.
- No screen recording. We do not record or upload any portion of your screen during a session.
- No AI-tutor claim. Study plans are transparent, rules-based, and tutor-approved - never presented as an automated verdict.