We have had an extensive exchange with refugees and they complained that the communication with German public officials is very tedious and fails at various language barriers. This concludes in the in the complete process being inefficient and frustrating.

What it does

Our Digital Companion ComYou fills various official documents automatically by conversing with the user and explains the implications of official response letters to the refugee in English and Arabic.

How we built it

We used Node.js in combination with NLU methods as a backend for our chatbot. After the user gave us all needed information, a request is being made to our second web server, which generates the application pdf based on the request body. This pdf then is stored in a blob storage and the link to access the pdf is send back to the user.

Challenges we ran into

Developing an NLP solution for the Arabic and Farsi Language. We could not evaluate the results very well because none of us speak either language.

Accomplishments that we're proud of

In less than 72h we researched the problem domain, conceptualized a solution and built the according prototype. We are convinced that our solution is solving a real-world problem.

What we learned

When building a Digital Companion it is essential to take cultural and language specialties into consideration.

What's next for ComYou

We are going to extend our solution with computer vision components and add more documents incrementally.

Built With

Share this project: