Privacy-First Family Device Management

01 / The problem

What needs to change

Families need useful device boundaries and access workflows without covert monitoring, unnecessary content collection, or controls that pretend to bypass platform safeguards.

02 / The approach

A focused path through the complexity

The product separates app-store-friendly family controls from supervised-device management, applies explicit policy logic, minimizes retained data, and limits AI assistance to privacy-preserving categorization.

03 / Key capabilities

What the system is designed to do

• Schedules, app and web rules, shields, and access requests
• Deterministic policy simulation and audit events
• AI-assisted URL categorization with raw logging disabled by default
• Supervised-device enrollment, profiles, restrictions, and command queues
• Retention, export, deletion, consent, and privacy settings

04 / Architecture

Technology and system shape

• TypeScript monorepo with shared domain and privacy packages
• Native parent, child, and consent-based desktop app scaffolds
• API, policy, filtering, worker, and supervised-device services
• Postgres and Redis deployment shape with local demo mode
• Admin console and generated client contract metadata

05 / Results

What the repository demonstrates

• The scaffold implements a working simulator demo, policy engine, API flows, administration, supervised-control previews, and privacy utilities.
• The design documents platform and entitlement limits instead of presenting unsupported controls as production-ready.

06 / Lessons

What the work teaches

• Trustworthy family technology begins with visible consent and data minimization, not surveillance depth.
• AI classification should be optional, bounded, and downstream of a deterministic policy model.

07 / Screenshots

Approved visuals