Full Size View

TECHNICAL

Active Stack
Node
Express
Mongo
Redis
Brain.js
Svelte
[ V1.0: ARCHITECTURE ]
Vinocat is a bidirectional decision engine for food and wine pairing, allowing users to start from a dish or a wine. It translates dishes into structured flavour profiles and aligns wines to the same model to generate ranked, explainable pairings based on compatibility and user preference signals.
[ V2.0: OBJECTIVES ]
Early-stage plans for v2 center on rebuilding Vinocat as a Go-based production system with a stable, deterministic pairing engine, while using Python as an experimentation pipeline to refine models, keeping the runtime clean, fast, and single-language, with ML improvements feeding in asynchronously.
April 2025 // v 0.0.1
// Genesis
Hardcoded pairing engine. Dataset limited to < 30 entities (Wine/Dish). Baseline validation of compatibility via < 12 basic pairing rules.
Vinocat Genesis v0.0.1
April 2026 // v 0.6.1
// Current State
Dataset contains 850+ entities (Wine, Beer, Dish, Cheese).

Dual-layer scoring engine: Objective compatibility via 56 granular, property-weighted heuristic rules; Subjective personalization via high-dimensional taste-vector association (Collaborative ML).
Vinocat Current v0.6.1