Sunshine Guardian

Inspiration 🌤

Increased levels and exposure to ultraviolet (UV) rays during the summer season pose a significant risk to skin issues and cancer. Despite the well-known importance of sunscreen application, only 30-40% of Canadian adults apply sunscreen. Many also fail to adhere to proper sunscreen application consistently due to forgetfulness and misinformation. Thus, Sunshine Guardian was created.

What it does ☀️

Sunshine Guardian is an app that reminds users to reapply sunscreen based on the UV levels of the user's location and the SPF of the sunscreen that the user uses. The app also provides sunscreen recommendations for users based on their skin type and complexion.

How we built it 🛠

We built Sunshine Guardian using a combination of different technologies:

• Figma for mobile app design and prototyping
• React Native/Expo for frontend development of our mobile app
• Python/Flask for backend development to handle API calls
• Python Libraries including Guardrails AI and Pydantic to produce structured LLM output
• APIs like Cohere, OpenWeatherMap, and OpenUV to populate app data

Challenges we ran into 🔗

While developing Sunshine Guardian we faced various challenges, including:

• Understanding how RESTful APIs work and how the frontend and backend ecosystems work together
• Setting up the new technologies (we tried using Flutter at first but there were a lot of system incompatibilities which meant we had to start again with Expo)

Accomplishments that we're proud of 🎊

• Using new technologies! Except for some basic Figma and Python, it was our first time using all these technologies and dealing with any sort of front or backend development. That came with a lot of reading documentation but also gratification and growth.
• Designing a user-friendly app that is practical and cute. We put a lot of thought and heart into the design to make it both aesthetically pleasing and accessible for our users.
• Completing our first hackathon! Not sure if the hunger or lack of sleep was worse, but the best part was definitely putting together a (semi) working product :))

What we learned 🌱

• To download software and check system requirements in advance
• Different ways to get structured output from interacting with an LLM
• There's almost an API for everything you would want

What's next for Sunshine Guardian 📌

• Developing more features like a hydration meter or leaderboard to gamify the app
• User authentication to save information across devices
• Improving the LLM recommendations using RAG and Cohere's custom connectors
• Deploying the backend with AWS/Heroku and looking into launching the app

Built With

• cohere
• expo.io
• figma
• flask
• guardrails
• javascript
• openuv
• openweathermap
• pydantic
• python
• react-native