Your private code is NEVER used to train foundational AI models without explicit opt-in.
Disable client telemetry with 1 line of JSON in your settings.
Full data subject protection under South African and international privacy regulations.
1. Overview & Scope
This Privacy Policy describes how Knolink (Pty) Ltd (“Entropy”, “we”, “our”, or “us”), registered in Midrand, Gauteng, South Africa, collects, uses, processes, and safeguards personal data and code telemetry when you interact with Entropy IDE, our website at entropy.knolink.co.za, and our cloud services.
By installing, downloading, or using Entropy IDE, you consent to the practices described in this Privacy Policy, incorporated into our Terms of Service.
2. Information We Collect
2.1. Account & Authentication Information
When you register an account or sign into the desktop or web IDE via OAuth (GitHub, Google, or Knolink SSO), we receive your primary email address, full name, avatar URL, and an associated account identifier.
2.2. Billing & Transaction Data
If you subscribe to a paid tier or purchase additional prompt credits, your payments are processed securely by our certified PCI-DSS Level 1 payment processors (e.g. Stripe, PayFast). Entropy does not store full credit card numbers or security CVV codes on our servers.
2.3. Customer Code & AI Prompts
When you invoke AI completions, chat, or agentic operations, the active code snippet or prompt is sent in transit to our Apolemia inference router or configured AI Provider (Anthropic, OpenAI, Google) solely to generate the requested Output.
3. AI & Zero-Retention Policy
Entropy guarantees that Customer Data, prompts, and codebase contexts are NEVER retained by default to train public machine learning models.
- Enterprise Privacy Agreements: Our agreements with Anthropic, OpenAI, and Google specify that data passed via API endpoints is not used to train or fine-tune public foundation models.
- Ephemeral In-Flight Processing: Prompts sent for inline autocomplete or agentic debugging are processed in memory and discarded upon completion of the inference session.
- Explicit Feedback Opt-In: If you choose to click a thumbs-up/thumbs-down rating on a specific AI response, you explicitly grant permission to analyze that specific snippet to diagnose model quality.
4. Telemetry & Opt-Out Controls
To improve IDE stability, track crash dumps, and optimize compiler latency, Entropy collects anonymized diagnostic telemetry (operating system version, memory footprint, startup time, crash call stacks).
You can completely disable local telemetry at any time in your settings.json file:
// settings.json
{
"telemetry": {
"diagnostics": false,
"crash_reports": false
}
}5. Subprocessors & Third-Party Sharing
We only share data with vetted subprocessors necessary to deliver the Entropy service:
6. Data Security & Retention
Entropy employs industry-standard security protocols:
- All API transmissions are encrypted via TLS 1.3.
- Database volumes and backups are encrypted at rest with AES-256.
- API keys and authentication secrets are stored in secure hardware-backed vaults.
7. Your Rights (POPIA, GDPR, CCPA)
Under South Africa’s Protection of Personal Information Act (POPIA), the European Union GDPR, and the California Consumer Privacy Act (CCPA), you have the right to:
8. Contact Information
For privacy inquiries, Data Subject Access Requests (DSAR), or compliance questions, please contact our Data Protection Officer at: