RideGo

• 

  Lisa wants to travel from Regensburg to Munich

• 

  But there are no rides available tomorrow at 6 pm

• 

  Matt also wants to travel from Regensburg to Munich tomorrow at 6 pm

• 

  It's a MATCH!!!!

• 

  Chat between Lisa and Matt to organize the trip

Inspiration

• You want to do a trip to another city
• Renting a car by yourself is expensive

• Share the rental in a seamless way

• Other ride sharing apps have unintuitive user interface

• We want to explore the possibilities Watson conversation offers for mobile development

What it does

Chat with IBM-Watson and tell it your trip details (date, time, start and destination). The backend searches for matches and connects the different parties.

How we built it

Client server architecture with IBM Watson (conversation), Python, Flask, MongoDB, Swift, Firebase. iOS app communicates with the Flask backend which forwards the messages to Watson and returns his results back to the client. After all needed details are passed the backend searches for possible matches in the database. It suggests to start a chat with the matched party to fix the details for the ride. The conversations between the parties are implemented with firebase.

Challenges we ran into

IBM Watson conversation: graphical programming and functionality Connect the mobile frontend with the backend and the Watson API

Accomplishments that we're proud of

Watson recognizes the needed infos (date, time, start, destination) and the trip matching

What we learned

To define conversations in IBM Watson and how to integrate it in a server-client-architecture

What's next for RideGo

• Increase the conversation flexibility
• Find car sharing partner(s)
• Improve app and backend

Built With

• chatbot
• firebase
• flask
• ibm-watson
• ios
• mongodb
• python
• swift