Inspiration
Organizational Practices, Optimizing Communication and Understanding
What it does
At the push of a button, automatically transcribes multiparty conversations, recognizing individual users and using AI and Cloud technology to extract insights and view them in an interactive tool.
How we built it
We combined numerous pretrained ML models from Azure with additional training for dynamic users and context.
Challenges we ran into
Learning new frameworks, deciding what features were important and what was practical to build in the time. Designing our interface to be user friendly as possibly.
Accomplishments that we're proud of
We had a number of moving parts, and did not start connecting them until afternoon of day 2. It was not until very late last night we even had a presentable demo. This was a big accomplishment.
What we learned
The power of pretrained models which, despite having some drawbacks, are extremely quick to set up and offer a surprising number of features.
What's next for WeMeet
We will meet again. Our code runs partially in the cloud and partially locally, but we would like to deploy the entire system onto a server so that we can show other people what we built.
Log in or sign up for Devpost to join the conversation.