FlavorFlow is an AI food recommendation engine built for food brands that want to move beyond generic marketing.
It ingests raw POS transaction data to construct rich, personalised taste profiles for each customer — capturing preferences across cuisine types, flavour profiles, dietary habits, and order frequency.
These profiles are embedded into a high-dimensional vector space using OpenAI embeddings and stored in Qdrant for lightning-fast semantic similarity search.
The system then uses these profiles to automatically generate targeted marketing campaigns tailored to each customer segment, ensuring every promotion feels relevant and personal.
Built with React.js on the frontend and PostgreSQL for structured data, FlavorFlow bridges the gap between transactional data and intelligent, AI-driven customer engagement.
Key Features
POS data ingestion and customer taste profile generation
Vector embeddings for semantic taste similarity matching
AI-generated personalised marketing campaign copy
Qdrant-powered vector search for real-time recommendations
Segmentation of customers by flavour preferences and habits