Building the Perfect Personalised Menu Using Python

Discover how a leading UK recipe box service utilizes Python to create personalized menus in this 21-minute conference talk from EuroPython 2020. Learn about the implementation of a menu planning optimization algorithm that ensures a diverse mix of recipes, dish types, cuisines, and ingredients. Explore the development of a recommendation engine that offers customers individually curated menus while optimizing operational performance. Gain insights into the use of Python packages like DEAP for genetic algorithms and integrations with graph databases such as neo4j. Delve into the methods, infrastructure, results, and lessons learned in building a personalization ecosystem for recipe selection and menu planning.

Introduction
About Gusto
Agenda
Overview
Deciding whats on the menu
Menu size
Planning menus
The problem
Datadriven menu
Genetic algorithm
Objectives
Minimising Unique Ingredients
Results
Personalization
Contentbased Recommendations
Collaborative Filtering
Personalization Results
Customer Feedback
Recipe Development
Lessons Learned
Outro