Meili

Meili is an application that creates a trip itinerary for users based on their preferences and historical data on what similar users have enjoyed.

Team

• Project Owner: Roman Gorelik
• Scrum Master: Erik Lin
• Team Members: Guillermo Adrian, Martin Glyer

Table of Contents

1. Team
2. Features
3. Recommendation System
4. Built With
5. Screenshots

Features

• Google and Facebook sign up and log in.
• Dynamic splash page allowing users to seamlessly transition around the page.
• Drag and drop activities or days.
• View up to date graphs representing user spending budget.
• View past and current trips in user profile, with ability to download PDF of the current trip.
• Quick onboarding process to allow better user recommendations.
• Recommendations change based on weather prediction.

Recommendation System

• Leveraged IBM Watson NLP and Microsoft Azure to categorize activities based on their description and images, respectively.
• Implemented Recombee, a lazy variant of rule-based recommmendation systems, to give user recommendations.
• Created weather prediction model using Prophet for weather-based recommendations.

Built With

• React
• C3/D3
• Semantic UI
• Express
• GraphQL
• Apollo
• Prisma
• PostgreSQL

Screenshots