↓ Skip to main content
Kanch’s Cloud Race Musings
  1. Personal Projects/

Learner Studio

·1 min
Table of Contents

Learner Studio
#

🎓 AI-Native, Artifact-Driven Competency Assessment & Mastery

Learner Studio is an AI-native learning engine designed to evaluate real-world builder capability. Learners submit authentic GitHub code repositories and receive an automated multi-agent diagnosis across an 8-dimension competency radar, targeted lesson modules, and closed-loop re-assessment cycles.

Originated from an ADK agent training initiative and generalized into a pluggable multi-domain learning platform.

How It Works:
#

Submit GitHub repo
    → Multi-agent analysis (AST parse + static scan)
    → 8-axis Competency Radar
    → Gaps identified → Learning modules assigned
    → Learner completes: Reading · Video · Whiteboard · Sandbox · Quiz
    → Re-submit → Radar expands → Repeat

The Tech Stack:
#

  • Frontend: Next.js 16 (App Router), TypeScript, Tailwind CSS v4
  • Design System: Sahara (warm linen & terracotta theme)
  • Visualizations: Recharts 8-axis radar charts
  • Analysis Engine: Python AST scoring & anti-pattern static scanners
  • Hosting Target: Google Cloud Run

Code: https://github.com/kanchpat/learner-studio/