Wandr - Chatbot

Inspiration

There is so much going on around cities, but many people don't put in the effort to explore, either because they're too busy or simply don't want to. Furthermore, it's easy to fall into routine and do the same things, eat at the same restaurants, etc. over and over. We hoped to create a tool that would give people a personalized list of things to do nearby.

What it does

Whether you're exploring your own neighborhood or visiting a new place, this bot aims to make you more aware of the goings-on in a given location. Based on a user's preferences, Wanda will make suggestions for nearby events, restaurants, and businesses that might be of interest to them. Users will be a able to set how often Wanda reaches out to them with new information (every 5 minutes, 30 minutes, couple hours, day, etc.) We hope this will help to enrich people's daily lives and increase community engagement.

How we built it

While we're still working on the product, we plan to use the facebook messenger platform to create a bot for our "Wandr" facebook page. With the user's permission, we will gather their location at an interval of their choice and periodically send out nearby events, restaurants, etc. We plan to obtain event and restaurant information using their lat / long info and the Yelp API.

Challenges we ran into

One of the major challenges was dealing with the boilerplate messenger code, which was for the gag clothing company "Original Coast Clothing". There was a lot more code and functionality than we needed for our MVP, so deciphering and manipulating the data flow was challenging. In fact, we're still working on this. Furthermore, we could only do testing on messenger on one computer, because the callback function, which redirected messages, took a single server URL which was either the deployed Heroku link or one person's local link.

Accomplishments that we're proud of

This was very challenging for us, and unfortunately our trajectory changed quite a bit in the earlier hours of the morning. If we had more time to familiarize ourselves with messenger, I think we would have been more successful and made a functioning app. However, I'm quite proud of what we were able to learn about messenger in this short time. We also had a bit of trouble making calls to Yelp's API due to authorization problems, but managed to work through this. I believe that these challenges made us stronger as a team, and this experience will leave us more prepared to succeed in high-pressure, time-intensive environments in the future.

What we learned

We learned about the challenges involved in working with an existing codebase. If we were to go back, we would have probably coded the messenger app from scratch, given the rudimentary functionality we were seeking for our MVP. We also learned that we could successfully grab location data from our browser window object and pass it to the backend.

What's next for Wandr

We hope to implement some kind of machine learning, so that Wanda learns about places and events you particularly like based on where you've been. Wanda will then suggest future events tailored even more to your interests. However, we'd also ensure that the bot throws in a unique event once in a while as well, because we encourage our users to explore and try new things.

Built With

• axios
• css3
• express.js
• html5
• javascript
• messenger
• yelpapi