Inspiration
This was our first time in Detroit, but my team was touched by the passion and pride Detroitians feel for their city. Despite rough times, detroit has an undying spirit, one that chicagoans can relate to. We decided to hack the commute after hearing about the unfortunate state of detroit public transit. One of the openning speakers told a story of a very bright young man who was unable to pursue an internship because the distance was unsurmountable. Thus came to our idea...
What it does
Our artificially intelligent chat bot, Poolio, enables users to request rides to a wide network of users around them, in the hopes that together they can share rides to work and build community. By doing so, both users contribute to bettering the city's traffic and improving the city's public transit.
How we built it
Poolio is is built upon the foundation of Microsoft's Bot Framework and LUIS cognitive services. The app is built in C# with 3 parts love and 4 parts blood sweat and tears. We worked in 6 hour sprints. By evaluating, brainstorming, individual and pair programming, we were able to iterate on our product and adapt quickly to a dynamic environment where new ingenious ideas consistently pop up.
Challenges we ran into
We initially worked with GitHub's Hubot, a tool for bot building. Our chat bot was coming along really well and was capable of a number of things including asking the user for their location and thus of determining their location vis-a-vis other users. However, working with Coffee Script for the backend work revealed to be a time-consuming and incredibly challenging task. So we decided to focus our attention to Microsoft's Botbuilder and its impressive machine-learning LUIS.
Accomplishments that we're proud of
Hackathon teams often times devolve into a work horse and some mules. One person may handle the bulk of the labor while their cohorts sleep, play and slack off. Our MHacks team worked consistently and cohesively to achieve our goal, and despite disparities in skill from one area to the next were able to make the most out of our time in Detroit. That is something we are all proud of.
What we learned
Personally, as this is my second time at a hackathon, I was quite unfamiliar with back-end development. Working with any APIs (whether Google Maps API, ArcGIS or LUIS) or with scripting languanges (JavaScript or Coffee Script), every moment was a wonderful opportunity to challenge myself to learn and grow in my knowlege.
What's next for Pool I/O
We plan to release our chat bot application to slack's integration directory. After which, any organization will be able to reap the benefits of integrated ride-sharing. We truly believe in the good that Pool I/O can do for the world, so we plan invest our time into improving the bot and moving the world closer to a shared economy!