[MAICE Dev Log 1] Building a Teaching AI, not just an Answer Bot (Project Overview)
How question-quality data, Bloom, and Dewey shaped MAICE's problem-definition strategy and multi-agent architecture.
#agent
Bongsam's Math & IT Blog
Recording stories about math, coding, utils, and education.
Sharing technical explorations for better education.
Recent Posts
View All →[MAICE Dev Log 2] How the agents collaborate (Multi-Agent System)
How MAICE split question classification, clarification, answer generation, and learning observation into a Redis Streams-based multi-agent flow.
#agent
[MAICE Dev Log 3] SvelteKit chatbot interface for real-time math rendering
How MAICE lowered the question-entry barrier with formula input, OCR, and streaming-safe LaTeX rendering.
#agent
[MAICE Dev Log 4] QAC checklist development: making educational quality measurable
How MAICE used QAC to evaluate question quality, answer quality, and learning-process support without mixing teacher and LLM scoring scales.
#agent
[MAICE Dev Log 5] Zero-downtime deployment for heavy AI containers (Blue-Green with Jenkins)
How Blue-Green deployment, health checks, and rollback scripts helped protect experimental continuity in MAICE.
#agent
[MAICE Dev Log 6] How we validated educational impact: thesis-based summary
A thesis-grounded summary of how question clarification affected dialogue quality, lower-quartile learners, and evaluation limits.
#agent#education#research