Data Governance & Protection
Privacy Policy
Effective Date: July 25, 2026 · Version 3.4
Privacy by Design Guarantee: KodeQualify enforces strict data minimization principles. We do not sell user or candidate data, train public LLM models on private code submissions, or store PII beyond necessary retention windows.
1. Information We Collect
We collect personal data to provide a secure, fraud-free, and high-fidelity candidate assessment experience. Information collected includes:
Account & Profile Data
Name, corporate email address, encrypted authentication tokens, workspace metadata, and billing info (processed securely via Stripe).
Candidate Technical Data
Submitted source code, execution stdout/stderr, keystroke velocity patterns, browser telemetry, and proctoring events during assessments.
2. How We Use Your Information
- Evaluating code execution correctness, test suite compliance, and runtime performance performance.
- Detecting anti-cheating signals, plagiarism across 7,500+ challenge banks, and synthetic AI assistance.
- Generating Scorecard Assist analytics for authorized hiring managers.
- Enforcing defensive HTTP Security Policies (`Permissions-Policy`, `Referrer-Policy: strict-origin-when-cross-origin`).
3. Subprocessors & Infrastructure Partners
We work with verified third-party vendors who adhere to SOC 2 and GDPR standards:
| Vendor / Entity | Role / Purpose | Data Location | Compliance |
|---|---|---|---|
| Amazon Web Services (AWS) | Cloud Infrastructure & Code Sandboxing | US / EU (Multi-region) | SOC 2 / ISO 27001 |
| Stripe Inc. | Payment Processing & Subscription Billing | United States | PCI-DSS Level 1 |
| PostHog Analytics | Privacy-Preserving Usage Telemetry | EU Cloud (Frankfurt) | GDPR Compliant |
4. Your Data Rights (GDPR / CCPA)
Regardless of your jurisdiction, KodeQualify guarantees the right to Access, Rectification, Export (Data Portability in JSON format), and Right to be Forgotten (Erasure).
Need to execute a Data Subject Access Request (DSAR)?
Contact our Data Protection Officer directly.