aul-ai-tutor

Autonomous offline AI tutor device for AI literacy in remote classrooms (local-first).

Roadmap: see ROADMAP.md

What it is

A local-only classroom device that runs an on-device LLM and delivers guided lessons without requiring internet connectivity.

Key capabilities

• Offline-first learning mode (no internet required)
• Local LLM runtime (on-device inference)
• Lesson engine + classroom safety controls
• Reproducible “golden image” builds for consistent deployments

Repository layout (initial)

• image/ – golden image build scripts and artifacts
• device/ – device services, configs, provisioning
• runtime/ – local LLM runtime wrappers / tooling
• curriculum/ – lesson packs and content (non-sensitive)
• tools/ – utilities (build, flash, validate)
• docs/ – architecture notes and operational guides

Quick start (placeholder)

1. Build image: tools/build-image.sh
2. Provision device: tools/provision.sh
3. Run tutor mode: device/start.sh

Roadmap

• Golden image v1 (reproducible build)
• Tutor mode MVP (lesson packs + safety)
• Fleet provisioning (offline updates, diagnostics)
• Hardware profiles and benchmarks

License

See LICENSE.