Planning meetings have always been a pain for us. Especially when it involves more than 4 persons. Doodle and emails works, but it's clearly not efficient and takes too much time.
What it does
Schedulo fetches everyone's calendar and find common availabilities when you talk about scheduling a meeting in a slack channel. It also recognize when someone says he can't on a particular moment to accommodate him.
How we built it
We developed a small flask app in python that connects the slack api and instantiate the real time messaging api for each teams that installs the app. When people chat in different channels, we send the messages to wit.ai natural language processing api to fetch the intent, subject, time and location of messages. With these informations, we can easily find availability for everyone to meet up without any effort.
Challenges we ran into
Wit.ai is a new technology. It is hard to use and to make sure results are consistent. We had to train the Ai with alot of use case, and even then, it needs a lot more training to become almost perfect. Also, being a team composed of developers made design and marketing way harder than it should have been.
Accomplishments that we're proud of
We are really proud to have successfully implemented a natural language processing api in less than 24 hours. We are also proud of the quick marketing plan we came up with even if we are not experienced in this kind of things.
What we learned
Developing an AI takes time and patience, it might not be the best thing to do in a 24 hours hackathon, but it sure was an interesting challenge.
What's next for Schedulo
We want to train it even more, so it becomes way more reliable. We are also looking forward to test it in real use cases.