El proyecto de fin de carrera q ue se presenta en este documento se ha realizado en una nueva empresa creada en Julio 2012 y basada en Paris que se llama SnapCar. SnapCar es una aplicación móvil conectada a una red de coches con chofer que van a buscar los clientes para llevarlos donde q uieren, un taxi moderno. Durante años , Paris tenía pocos taxis y se necesitaban más, y aún hoy , la situación es la misma, y con más flexibilidad y mejor calidad de servicio. SnapCar es una solución a esos problemas. La propuesta de valor de SnapCar es pon er en relación el cliente con el coche más cercano. Es posible gracias a las nuevas tecnologías móviles: geo - posicionamiento y contratos de datos nos dan informaciones sobre el sitio donde está el usuario, la posibilidad de conectarlo directamente con el c hofer y soluciones para seguir el camino del coche en directo. Esas posibilidades han cambiado toda la experiencia del taxi. El canal principal por el cual SnapCar interactúa con sus clientes es la aplicación móvil. SnapCar ha desarrollado todas sus tecn ologías con gente suya: web server, back - office y la aplicación iPhone. Pero el equipo inicial tenía limitaciones en tiempo como en recursos humanos y por eso encargó el desarrollo de una versión Android a un contratado de una empresa externa. Esa versión fue la 1.0. Ofrece ya las funcionalidades básicas de SnapCar. Pero cómo la versión por iOS estaba desarrollada internamente, ha recibido al cabo del tiempo nuevas funcionalidades y mejoras. La necesidad de mejorar también su aplicación Android es lo que ha creado esa práctica. English: The master thesis exposed in this document was realized inside a young company, a start-up called SnapCar launched in July 2012. SnapCar is a mobile application with a network of chauffeured vehicles that will pick clients up on demand, a modern day's taxis. Paris has suffered for decades from a weak taxi offer: it needed more flexibility, quality and quantity. SnapCar is a solution. The core business of SnapCar is to provide the connexion between a client and the nearest car. This solution is mainly based on mobile solutions: geo localization and data exchanges provide data about the customer whereabouts, a way to connect him directly with the incoming driver and a mean for the client to trace in live the vehicle's course. These possibilities are changing everything in the way people experience taxi rides. The only connection with the client is the mobile application. SnapCar has developed most of his technologies internally: server, back-office, iPhone application. Limited on time and human resources, it outsourced the Android application development. This version, numbered v1.0 provides the core functionalities SnapCar wants to offer as services. But it lacks several others ones and presents weaknesses. The iPhone application was developed internally and thus received updates and new features. There was a real need to improve the Android application as soon as possible, based on the previous code, and it creates that internship opportunity. The logical set-up of screens and user information had to be reworked in order to have a more efficient access to important data and more readability. SnapCar application provides two ways to book a car and both needed a rework for better performances, logic and display. A whole set of feature had to be added in order to offer business solutions. Several details here and there had to be added too. Other issues of this 1.0 version are some instabilities issues with recurring bugs causing crashes. Another group of improvements were needed to keep up with the fast pace of the mobile technologies. The implementation of the tabulation views using fragments could be questioned as the only-one-activity architecture. To resume, the version 1.0 of the Android application presented a wide set of issues that needed to be prioritized and addressed: the aim of this master thesis was to develop the application to offer the full services of SnapCar but also to investigate and eventually implement better solutions to tackle some of the broader challenges. The project was supervised by the lead engineer and by the founder. It was done in several steps with some parallelism: to discover and get knowledge of the version 1.0 code, to focus on functionalities, to publish an intermediate version numbered 1.1, to investigate on better solutions, to implement these solutions toward a version 2.0. The team uses RedMine to structure its developing efforts. This tool allows to create tickets, considers versions and tasks' percentage of completion. Being a start-up with only 2 engineers, the development progress is instantly communicated. At the time of writing, the project has completed most of his objectives and only the last one, the implementation of the selected enhancing solutions is still in progress.