WanderWith

splash screen

Inspiration

*Imagine you've got some time in a new city, after an interview or meeting. You want to go around and explore the town a bit, or do some of the really cool local stuff. More than often, you might just choose to stay in your hotel because it can be boring to do things alone. What about an app that helps you find people to wander with, people who are interested in the same thing? No planning, no scheduling nothing, no strings. Impromptu, just the way a nice fun outing should be. *

What it does

The app perfectly captures the idea of having less time in hand when we visit a place for a short stay and having enough of time to get bored at the hotel. So, why not have some one to explore the vicinity around you and with you. The app takes the activities and places of interest from all the users who have checked in to the place during the last 24 hours. The idea of the app is to suggest travel buddies to the users on the fly within minimum clicks. As the app doesn't require any prior planning about the tour, it becomes really easy for the user to get 'WanderWith' buddies. The app incorporates login with Facebook to maintain the authenticity of the users and in future, Facebook profile information can be used to know about the interests of the user so that we can pro actively suggest potential places to the user with new check in.

How I built it

We began by experimenting with Amazon and Google's cloud platforms, and managed to set up a Compute Engine as well as a CloudSQL instance, but later realized that it would be better idea to set up a basic development server first, after which we moved to our local TAMU servers. We used the AJAX + Django stack to build our webapp, as many users tend to prefer websites over apps, to avoid over-cluttering their mobile phones. We had a couple of members working on Bootstrap to build a interactive UI, as well as Javascript code to communicate with our Python backend. We used Facebook for authentication, and also as a means of identification. The other members focused on writing backend code to execute recommendation algorithms and database interfacing.

Challenges I ran into

We faced numerous challenges while coming up with the basic prototype of the app idea. We tried to deploy our application on Google Cloud and make a domain name on MLH but the deployment on google cloud took unnecessary time and the MLH domain name didn't get activated till now. We began by having a vision of very intense app idea but it turned out that we had a very tight time frame and we ended up implementing the minimum viable application that can perfectly capture our idea.

Accomplishments that I'm proud of

We pitched our idea to many of our co-hackers, and they all agreed that there is immense scope to solve such a problem, and monetize it over the long run. Most of us have faced this problem, especially when going to a new city for an internship, on-site interview or meeting in cities such as New York, Seattle and Austin. While most current apps attempt to unite travelers via prior planning and travel coordination, our app tries to retain the charm of impromptu planning by allowing the users to search for 'co-wanderers' as and when they feel like it, and find time in a new city. It is hard to make an app that depends on such simplistic user input and interaction, and yet can solve a very crucial and challenging problem faced by many all over the world. We believe we have been able to come up with such a minimalistic design that can capture just enough data to link up the user with potential travel partners.

What I learned

How to convert a raw idea into a minimum viable product, and come up with minimalistic design that will help us pitch it.
The usage of cloud services and various new web APIs, which are powerful beyond imagination, and need to be handled carefully, simply because there are so many of them.
Dealing with the trade-off between fixing minor bugs and pushing to add more features in the short time span.

What's next for WanderWith

WanderWith has a tremendous scope to be scaled up to the level where the suggestions can be made using Machine Learning techniques like Collaborative Filtering, Matrix Factorization and Recommendation System Models. WanderWith can further make use of Facebook Graph API and give preferences to local experts and friends while making suggestions.