Awair

Inspiration

Credits to Ryan Brigden for coming up with the idea. Typically, public safety news arrives to us via journalists. Due to this, there is generally a long delay between when events happen and when the public discovers it. Our goal is to eliminate this delay.

What it does

Transcribes and disseminates public safety transmissions from Broadcastify and uses Microsoft's Project Oxford speech API to transcribe the radio in real time.

How we built it

-Parses Broadcastify audio by long periods of silence using pydub
-Transcribes audio files using Project Oxford speech recognition
-Creates JSON text files of the transcribed audio segments
-Ran our server in an Azure VM
-JSON objects stored in a database
-Website displays alerts

Challenges we ran into

-Broadcastify audio parsing
-Server issues
-Webdev

What's next for Awair

We'll expand it to work in major cities across the United States.

Built With

• ruby-on-rails
• python
• microsoft-project-oxford
• heroku
• broadcastify
• pydub
• azure
• ubuntu
• json