mhoover/mealtime

Analytics and architecture for a web app predicting food delivery times in Nairobi, Kenya # MealTime MealTime is a project built for the Insight Data Science New York City summer session from June through July 2015. The purpose of MealTime is to predict more accurate delivery times for online food delivery in Nairobi, Kenya. Analytic work utilizes Python's scientific suite of packages and uses proprietary delivery data from a food delivery service in Nairobi as well as Weather Underground's API for associated rain data. Prediction times are based on a gradient boosting model. More details on the analytics involved can be found here. Front-end work was done using the Flask, a Python-based web framework, with Twitter Bootstrap themes. All work is hosted on Amazon web services. For more information, contact matthew.a.hoover at gmail.com.