Frequently, speaking with someone who is a "mentor" can be a daunting task. Trust is hard to build with authority figures, and a lot of times it's too scary for young people to follow through. We wanted to help break down the barrier of finding help by evening out the playing field. We wanted to help bridge the gap between students that can easily slip through the cracks (under-privileged students, perhaps at a magnet school) and students that have experienced similar things at a different institution to talk about common problems that they are both facing.

What it does

Peer to Peer is a platform for youth to connect to youth mentors. Anyone can get help, any time. You can either choose to give help, or receive help. This is done via queries into a search field stating your particular problem. After we have received your problem, we match you with another user that has selected to give help in that area. You then are put in a chat room with your peer to discuss about moving forward and getting the proper help. You can speak your problem to the platform (speech to text) or you can upload a picture of a problem to get relevant keywords and search that way.

How we built it

Peer to Peer is built on javascript, html, and materialize css. We also use Microsoft Cognitive Services API to handle image to text. Speech to text is handled by Speech io. Ngrok is being used for the backend server stuff (Dylan is a beast and a half for setting this up).

Challenges we ran into

Peer to Peer started off as a React project, however, after several failed attempts at initializing projects and npm issues, we decided to pivot to something we all knew (javascript/html/css3). Gina noted troubles in configuring and using the speech to text api.

Accomplishments that we're proud of

Functioning chat rooms, speech to text, easy navigable UI/UX, image to text, completing a hack.

What we learned

It's really hard to set up React...

What's next for Peer to Peer

Add more functionality like more subject areas to get help on, move to a framework, create robust matching algorithms...

Built With

Share this project: