Inspiration

A few weeks ago one of our members had a scare with his grandmother. Her blood pressure went up and this caused her vertigo, which could have caused a stroke or a fall that could have been fatal.

What it does

Granny's Watch is a webApp, based on a voice message, which can detect when a person is in danger. Once detected, a marker is placed on a map, so that someone nearby can provide medical support until the ambulance arrives. In addition, a message is sent to people nearby, children or relatives, to go running.

How we built it

We have used HuggingFace, Cohere, Pandas, Matplotlib, Streamlit and Flask to perform the product and analysis.

Challenges we ran into

We participate in the EDA 3rd challenge and the best use of Cohere in NLP.

Accomplishments that we're proud of

We are proud to be able to realize a creative solution with the multidisciplinary team. Each one has had a clear role and this has allowed us to move forward and finalize a first value proposition.

What we learned

We have also been able to learn new libraries such as HuggingFace, Cohere and Streamlit.

What's next for Granny's Watch

In the next iterations it would be interesting to be able to realize the protype instead of a WebApp on a Smart Clock, so that the user experience is as fast as possible. Since in the elderly a few minutes can be fatal for their life.

Built With

Share this project:

Updates