如何以 FalkorDB 和 Streamlit 建立圖譜驅動的學習智能體 Campus Brain

How I Built a Graph-Powered Study Agent Using FalkorDB and Streamlit

Introduction

During the Graph Hacks hackathon, I wanted to tackle a common problem in education: navigating complex curriculum prerequisites and tracking personalized study progress. To solve this, I built Campus Brain—a graph-powered study agent designed for Data Structures.

The Architecture & Tech Stack

To make the application fast, flexible, and capable of maintaining state, I used a modern stack:

• FalkorDB Cloud: Used as the core graph database to store educational dependency graphs (PREREQUISITE_OF relationships) and manage private student memory (MASTERED states).
• Streamlit: Used for the interactive user interface, allowing students to select topics, track their progress, and view live updates.

FalkorDB in Action & Cypher Queries

Graph databases are uniquely suited for handling interconnected data like course prerequisites. Instead of running heavy relational joins, Campus Brain uses Cypher queries to dynamically traverse the curriculum graph.

For instance, finding the prerequisite chain for a target topic looks like this:

cypher
MATCH p = (a:Topic)-[:PREREQUISITE_OF*1..8]->(t:Topic {name: $target}) 
RETURN a.name, a.unit, max(length(p))

Additionally, FalkorDB powers the persistent memory feature. When a student marks topics as studied, the app records the state in a private student graph session so the agent remembers their mastery across sessions.

Conclusion
Building Campus Brain with FalkorDB was an incredible experience. The cloud setup was seamless, and combining graph queries with an interactive frontend opened up powerful possibilities for AI agent memory and structured learning paths!