Machine Learning/Reinforcement Learning Expert (Python) to work on the first phase of a mobile app for taxi sharing (only Python required)

This first phase is really simple and there's a GitHub code repo (see below) that can be built upon. Simply train a model (using DQN) that will pick two people from an origin and decide whether they can share based on their desired arrival times (e.g. if both are going to same destination and have same arrival time, then they can share. If they're going to separate destination and one is going in the morning and the other in the evening, then they have to be taken separately.
This GitHub code repo can be modified for just one car and two passengers (assume passengers have same origin address):

https://github.com/donjpierce/traffic