Home/Legal/Privacy Policy
๐Ÿ›ก๏ธ
Accurate Data Disclosures

Privacy Policy

Comprehensive, honest disclosures detailing exact data collection, in-browser storage, transient audio processing, and zero cloud audio retention practices.

Summary of Provisions:Your classroom audio stays in your browser and is never permanently stored on cloud servers. We practice ephemeral processing, zero audio retention, and client-side data ownership.
Entity: Zoef TechยทRegistration: MSME and Udyam Registered EnterpriseยทJurisdiction: Republic of IndiaยทLast Updated: November 2026

1. Data Protection Commitment & Controller Identification

Zoef Tech ('we', 'us', 'our'), an MSME and Udyam registered enterprise in India, is dedicated to the highest standards of privacy, confidentiality, and data minimization in education.

This Privacy Policy outlines precisely how AI Class Observer collects, processes, and protects your information, fully compliant with the Digital Personal Data Protection Act, 2023 (DPDPA, India) and the Information Technology (Reasonable Security Practices and Procedures and Sensitive Personal Data or Information) Rules, 2011.

2. Technical Architecture & Actual Data Flows

Our software is engineered with a strict 'Privacy First' architecture. The following represents the actual technical data handling in our codebase:

  • Audio Capture & Local IndexedDB Persistence: When you record a classroom session, raw audio chunks (1-second slices) are stored exclusively inside your browser's local IndexedDB database ('cor_audio_records'). This ensures crash-resilience in case your device runs out of battery or the tab refreshes. At this stage, NO data has left your device.
  • Transient Secure Transmission: When you press 'Generate Report', your audio is transmitted over an encrypted TLS connection (HTTPS) to our backend processing endpoint (`POST /api/reports`).
  • Ephemeral Speech-to-Text Transcription: The backend streams the audio to the Groq Whisper API (`whisper-large-v3-turbo`) for transcription. The audio stream is processed in transient memory.
  • ZERO Permanent Cloud Audio Storage: We DO NOT store, archive, or retain your raw classroom audio on any server disk or cloud database. Once speech transcription is complete, the audio buffer is wiped from runtime memory immediately.
  • Pedagogical Analysis: The textual transcript and user-provided session metadata are evaluated by high-speed inference LLMs (Groq / Gemini / OpenAI / DeepSeek) strictly to calculate teacher/student talk ratios, detect pedagogical commendations, and generate rubric evaluations.
  • Tamper-Evident Registry Record (MongoDB): The generated report JSON (metadata, metrics, recommendations, and a unique 24-character hex ID) is saved in our secure registry database to allow instant online verification (`/verify/[id]`) and QR code validation. NO audio is ever saved in MongoDB.
  • Local Report Ownership (localStorage): The final PDF report and local history are saved directly in your browser's `localStorage` ('cor_observation_history'). You have complete 1-click control to download, inspect, or delete records from your device.

3. Information We Collect

We collect only the bare minimum data required to deliver the observation report:

  • Session Metadata: Teacher Name, Grade / Class, Subject, Lesson Topic, and Observation Date.
  • Classroom Audio & Spoken Content: Processed transiently during session generation, never retained or used to train public machine learning models.
  • Technical Diagnostics: Browser type, device screen category, and error codes necessary to maintain application reliability.

4. Student Privacy & Non-Biometric Processing

AI Class Observer is designed specifically for educator self-reflection. The system does NOT create voiceprints, voice biometric templates, facial recognition models, or student profiling profiles.

Classroom acoustics are analyzed strictly at an aggregated linguistic and acoustic level to distinguish teacher talk patterns from collective student participation percentages.

5. Cookies and Client-Side Storage

AI Class Observer does NOT use third-party advertising cookies, behavioral tracking pixels, or cross-site tracking scripts.

We use only essential client-side storage technologies (HTML5 LocalStorage, IndexedDB, and Service Worker caches) strictly necessary to enable offline Progressive Web App (PWA) functionality, report history viewing, and crash-resilient recording.

6. User Rights and Data Deletion

Under the Digital Personal Data Protection Act, 2023, you have the right to access, review, and delete your personal information. You can:

  • Delete any saved session from your browser instantly using the 'Delete' button on individual report cards.
  • Purge all local reports and IndexedDB audio cache immediately using the 'Clear All History' action in the Reports tab.
  • Request removal of a specific 24-character Registry ID from our online verification database by emailing privacy@zoeftech.com with your report ID.

7. Grievance Redressal Officer

In compliance with the Information Technology Act, 2000 and DPDPA, 2023, the contact details of the Grievance Officer for privacy and data protection concerns are:

Grievance Officer: Zoef Tech Legal & Privacy Department Entity: Zoef Tech (MSME & Udyam Registered) Email: privacy@zoeftech.com / legal@zoeftech.com Jurisdiction: Varanasi, Uttar Pradesh, Republic of India