03 Dec Case Study: Uber Bot
Following the tremendous chat bot market growth, there was an obvious need to conduct research in this field and gain chat bot experience in Python, NodeJS, Azure and MS Bot Framework. With this in mind, an effort was taken to develop a chat bot for the Uber ride-share service.
Market analysis showed that there was no working bot for requesting Uber rides through popular messenger applications. Our team saw an opportunity to create a bot that would allow users to request rides without having to log into the Uber applications itself. This way, users could request rides from a desktop computer, through a messenger application, or through their phone without having to log into the Uber app. The chat bot would even allow users to start a request from a desktop and finalizing the request from their phone. The first version of this application was developed using Python and was available only via Telegram. After some testing, decision were made to develop a new version from scratch, the second version was based on Node.js and TypeScript stack and used MongoDB for data storage while utilizing Microsoft Bot Framework + Azure.
This technology allowed integration between the bot and Telegram and Facebook Messenger. Messengers like Slack, MS Teams, Skype, GroupMe, Kik, Skype for Business can be supported. Also, it is now possible to embed the bot into web pages, email, SMS and even Cortana. In order to attract users to the bot, a small marketing campaign was held (bot was listed in several rankings) and a landing page was developed.
Benefits and Results
The bot was released to the public and initially attracted around 100 new users monthly, with several rides requested each day by users in different countries. The major benefit is the gained practical experience in working with Microsoft Bot Framework, and experience in development and promotion of chat bots.
Python, MS Bot Framework, Azure, Node.js, TypeScript, MongoDB