#Security #DataTier

Effective date: August 1, 2026

7.1 Purpose

To define consistent sensitivity levels for all data handled by Hirfa, so that appropriate, proportionate protection controls are applied to each category.

7.2 Classification Levels

  • Level 1 — Public: Data intended for public consumption: marketing website content, public app-store listings, promotional campaign assets. Controls: No special handling required; standard publishing review only.
  • Level 2 — Internal: Non-public operational data: internal product roadmaps, performance dashboards, internal meeting notes. Controls: Accessible only to Hirfa Personnel; must not be shared externally without authorization.
  • Level 3 — Confidential: Personal data of Users: names, phone numbers, addresses, booking history, in-app chat content, ratings and reviews, Craftsman trade credentials. Controls: Encryption in transit and at rest; role-based access restricted to Personnel with a legitimate operational need; access is logged (see Section 12).
  • Level 4 — Restricted: The most sensitive data categories: national ID numbers, payment tokens/financial account details, authentication secrets (hashed passwords, OTP codes), and security incident evidence. Controls: Highest protection tier — access limited to a minimal, named set of Personnel; dual-approval required for any exceptional access request; full audit logging of every access event; encrypted at rest with restricted key management.

7.3 Handling Rules by Level

  • Level 3 and Level 4 data must never be stored on unmanaged personal devices or transmitted via unencrypted channels (e.g., plain email, consumer messaging apps).
  • Any transfer of Level 4 data must occur exclusively through approved, encrypted internal channels, with prior authorization from the designated security owner.
  • Level 3 and 4 data must be excluded from non-production (development/testing/staging) environments unless irreversibly anonymized or synthetically generated.

7.4 Periodic Review

Data classification assignments are reviewed at least every six months, and immediately whenever a new feature introduces a new category of sensitive data.