The problem TalkShort solves

In our day to day lives there are many scenarios where we’d like to jot things down else we will definitely forget it.
It’d be nice if we could record these conversations and have them stored in a text format for future reference and recollecting the small details which we could have forgotten.
And what’s even better, have the text summarized. It saves your time and gives a concise gist of the information
Use Cases:
An elderly man goes to a doctor, the doctor might prescribe various medicines to him, and bunch of medical advice. It could be a lot for anyone to process it.
Seminars are really interesting events with an overwhelming pace of information being thrown at you. This app could save us noting things down quickly and enjoy the talk.

Challenges we ran into

For text summarization, we had to try hard to make it an accurate model. We tried a few other algorithms but they would have needed training time and resources which we were devoid of, in those circumstances we chose the most efficient algorithm available to us that doesn’t need training. Some of them did'nt provide concise and ligible summarization.

We were making an android app using java, we used java for most of the functionality. For the text summarization we wrote a program using TextRank algorithm in Python. It was tricky to integrate python with java. So we wrote the code in java itself by keeping it equally efficient.