Back to Projects
Hackathon III: Learnflow AI Platform
Hackathon Project

Hackathon III: Learnflow AI Platform

A production-grade AI tutoring platform built with Reusable Intelligence and Cloud-Native agentic infrastructure.

Live Demo Source Code

The Challenge

Shift from syntax writer to system architect by teaching AI agents how to build sophisticated cloud-native applications using 'Skills' with MCP Code Execution.

What I Built

Constructed 'LearnFlow', a Python tutoring portal. Built a library of reusable agent 'Skills' that allow Claude Code and Goose to autonomously manage the entire stack.

Key Features

Reusable 'Skills' library for agent autonomy
Token-efficient MCP Code Execution pattern
Multi-agent tutoring orchestration (Triage, Concepts, Debug)
Struggle detection triggers and automated mastery tracking
GitOps delivery using Argo CD and GitHub Actions

Tech Stack

Next.js + MonacoFastAPIDaprKafka on K8sArgo CDKong GatewayMCP Code ExecutionGoose

Achievements

"Achieved 90%+ token reduction using advanced execution patterns and established a library of portable AI skills for autonomous infrastructure."