FINDORA — Graph-Based Social Recommendation

Overview

I built a socially-supported recommendation platform for films, series, books and music, modeling user–content relations in Neo4j and designing a multi-layer recommendation engine with Cypher.

Technical Approach

• User–content relationships modeled in a Neo4j graph database.
• Dynamic recommendations produced with Cypher queries and graph traversal.
• FastAPI RESTful API managing the system's entire data flow.

Key Features

• Scalable backend architecture processing user interactions such as likes and follows
• Modern, responsive Flutter interface
• API-based dynamic structure for real-time interactions and data updates

System Architecture