Blaze Buddy

💫 Inspiration

There is a growing issue of wildfires happening across the globe. Due to the rise in temperatures over the past few years, extreme heat waves have become more frequent, making it 5 times more likely for wildfires to occur. Unfortunately, fires have caused numerous deaths yearly. For instance, in 2021, over a million fires claimed the lives of around 3.8k people and injured more than 4 times that amount. This led us to ask ourselves, 'What can we do to keep people safe during fire incidents in such conditions?'

😮 What it does

Our Application, Blaze Buddy aims to provide people with accurate and real-time information about fires in their area to prevent any delay in response. Blaze Buddy includes information on the news covered and uncovered fire incidents, as well as other information such as fire safety protocol. Our goal for Blaze Buddy is to be a one-pit stop for users on staying informed about fire-related incidents, statistics, and safety protocols.

🔨 How we built it

We built Blaze Buddy's Machine-Learning algorithms using Cohere.ai in Python. We used NASA FIRMS API to get real-time satellite data on fire incidents across North America, and we used Cohere generate model to create a summary of the incident. We then used NewsApi to get recent news articles, we used Cohere embed model to identify fire-specific articles, as well as Entity Recognition to scrape the location of the incident.

Our front end was first brainstormed through Figma, and afterward built with React.js and Material UI, and we Flask to connect the data we collected on our back end to our front end.

😰 Challenges we ran into

The first challenge we ran into was connecting our back end to our front end as we spent a lot of time developing our algorithms, so we did not get much time to focus on getting our data onto our website.

😤 Accomplishments that we're proud of

We are proud of using NLP models, along with APIs to scrape relevant and meaningful data for our use case. We are also proud of being able to develop a product with new technology that we haven't used in the past in the time frame given.

🧠 What we learned

Research counterpart We learned a lot about Cohere.ai and how to use their models to help us bring our idea to life. We learned to utilize Cohere NLP models, along with APIs to scrape relevant data to provide to our users.

Web Development counterpart None of our group members had any experience with Flask so this hackathon was a great learning experience in learning a bit more about connecting a back-end to a front-end, allowing us to build a full-stack application.

🔥 What's next for Blaze Buddy

We have a list of things we'd love to do for Blaze Buddy in the future!

To improve on accessibility features (e.g colour blindness, motor disabilities)
To get real-time data around the world
to add a chatbot (we believe its essential to people who want to learn more about fire safety)