A production-grade AI guide system for Chinese intangible cultural heritage museums, combining knowledge graph RAG, community UGC, and a consumption loop with multi-agent orchestration and rigorous quality assurance.

Stars

112

7-day growth

No data

Forks

4

Open issues

0

License

No data

Last updated

2026-07-27

AI repository intelligence
FR-AI / ANALYSIS

Why it is worth attention

It stands out for its comprehensive, engineering-first approach: detailed decisions on word mapping, context compression, and a multi-sink self-correcting review agent, all backed by measurable KPIs and a full visitor-to-operator closed loop.

Who it is for

  • General museum visitors seeking deeper understanding
  • Enthusiasts and collectors wanting artifact lineage and cross-category comparisons
  • Artisans and content creators aiming to share knowledge and build influence
  • Museum operations teams needing data-driven insights and content quality monitoring

Use cases

  • Real-time guided tours with GPS location and AI narration adapted to user interest
  • Multi-hop knowledge graph queries for artifact history, craftsmen networks, and cross-category associations
  • Community UGC with tiered creator levels, achievements, and AI cold-start content generation
  • Operations dashboards displaying KPIs, hallucination alerts, visitor profiling, and A/B testing

Strengths

  • Hybrid retrieval (pgvector + FTS + RRF + word mapping) achieves HitRate@3 ≥85% and Faithfulness ≥0.85
  • ReviewAgent implements a five-sink asynchronous feedback loop (event bus, lesson store, audit queue, etc.) for production-quality self-improvement
  • Personalized user experience through cognitive-level narration, interest profiling, and multi-modal input (photo, voice)
  • Complete value chain from free knowledge dissemination to paid courses, merchandise, and community sharing, driving repeat visits (≥25%) and conversion (≥10%)

Considerations

  • GPS accuracy is hall-level (~5-10m), requiring manual 'I’m here' fallback for precise positioning
  • Current knowledge graph handles <10,000 entities; scaling beyond 100,000 may require migrating from PostgreSQL CTE to Neo4j
  • Complex multi-agent architecture and review pipelines increase operational overhead and dependency on multiple infrastructure components

README quick start

快速开始

Related repositories

Similar projects matched by category, topics, and programming language.

lopopolo
Featured
lopopolo GitHub avatar

harness-engineering

Harness Engineering is a methodology for improving coding agent outputs by carefully crafting the environment around them—providing curated context, tools, and executable constraints that encode an organization’s nonfunctional requirements and cumulative lessons.

AI & Machine LearningAI Agents
2,390
slvDev
Featured
slvDev GitHub avatar

esp32-ai

A 28.9 million parameter language model runs on an $8 ESP32-S3 microcontroller entirely on-device, generating simple stories at about 9.5 tokens per second.

AI & Machine LearningLarge Language Models
1,960
littledivy
Featured
littledivy GitHub avatar

mimic

mimic captures traffic from any iOS or web app and automatically generates a Python client library that lets you call the app's API like a regular library.

AI & Machine Learning
1,482