As current college students living in the city without a car, it is vital to take one of these transportation services to get around. Since we are college students, we are trying to save as much money as possible and are always looking for the least expensive products and services. Travel Time Tweets is the solution to this problem many college students face getting around.

What it does

This artificially intelligent twitter bot recognizes its mentions and retrieves geolocations from the respective tweets and automatically Direct Messages the user that mentioned it. The bot asks the user for their destination and instantly returns a comparison list of all possible transportation methods from least expensive to most expensive and then allows the user to request a ride. The bot compares prices by using asynchronous parallel programming.

How I built it

We split up in two teams, one tackling the twitter bot itself, and the other dealing with the transportation APIs and creating a comparison list. The packages we used were: Twit, Geocoder, Node-Lyft, Node-Uber, and the languages used were: Node.JS, JavaScript, JSON, Git, as well as Bash. We then deployed the twitter bot with in a client-server relationship with the cloud Heroku.

Challenges I ran into

There were several challenges that we encountered while developing our project. One of the most significant challenges was our collective inexperience of APIs and Node.js, which is what our whole project was developed from. After reading documentation pages, we then encountered a few bugs through the development process. Fortunately, we utilized the iterative software development process and were able to code small sub-tasks and fix these bugs before merging all of our work. One of the notable sub-task issue was our lack of knowledge to modify the permissions of the Twitter account to access direct messages, a vital part of our project. Since we made several web requests to servers, another sub-task issue was staging callback functions while multi-threading.

Accomplishments that I'm proud of

We are proud that we developed a functioning Twitter bot that can read and write tweets, follow users, and have artificially intelligent conversations through direct messages.

What I learned

We learned how to code with Node.js and use APIs in this 24 hour experience with no previous background knowledge. We also learned how to use Git in a collaborative project and be able to merge many, many conflicts between versions of the same file. We also learned a lot about JSON files and manipulating them to parse into arrays and strings.

What's next for Philly Codefest 2018

We plan to continue adding more features, making this nationwide accessible, and hopefully getting all college students to start using Travel Time Tweets to assist in transportation endeavors.

Built With

Share this project: