FlareTracker

Inspiration

I have multiple chronic skin conditions that often flare up without warning. Over the years, I’ve tried to track my symptoms, medication usage, and possible triggers using paper notes and various apps, but I could never stick with it consistently. Most tools I've used felt too clinical, too rigid, or just didn’t fit into my routine.

Chronic skin conditions are notoriously hard to monitor; symptoms can change daily, progress is hard to see, and it’s often difficult to remember what’s helped (or made things worse). I built FlareTracker because I needed something smarter and simpler, something that could help me understand my own skin better, and hopefully make the day-to-day a little easier.

I've had real success talking with ChatGPT (it's been loads better than any dermatologist I've seen) so I wanted to integrate conversational AI into the app too!

What it does

FlareTracker is a skin health tracking tool designed to help people track chronic skin conditions like eczema, rosacea and acne. It allows users to add a check-in to their daily routine, logging the severity of their conditions, symptoms medication as well as other lifestyle factors like stress levels, diet and sleep quality which may have an impact on their skin.

Users then have the ability to explore their conditions over time, correlate medication usage or lifestyle factors, even possible flare up triggers. FlareTracker builds up a detailed picture of the users conditions over time. This can be shared with the users medical professional.

A conversational AI assistant has been added that allows the user to ask specific questions about their conditions, medication usage, lifestyle factors as well as any patterns that might be forming.

How we built it

I used bolt.new of course!!! The basic app was completed in less than one day, I continued adding features over the month. I had to implement some of the fixes myself, but almost all of the work was completed within the bolt.new code editor.

I was able to use ElevenLabs text-to-speech for the voiceover in my video, I also relied on ChatGPT for the video script and logo design.

Challenges we ran into

Bolt is an amazing tool, it really is. However there are a few things that I kept asking it that it just would not fix (eg: you can probably see in the video the modal overlay top margin isn't quite right. Bolt kept telling me it had fixed it but it had not).

Apart from that, the biggest challenge was rationing my tokens and getting my prompts just right!

Accomplishments that we're proud of

Honestly I'm proud that I completed the project AND the video, I am a serial 'starter' and not a great finisher when it comes to development.

I'm really happy with the final result and I'm really hoping this tool will help others as well as myself!

What we learned

I learned quite a bit about prompt creation, UI/UX as well as integration with Supabase. I discovered some awesome new tools (ElevenLabs, Tella, Canva) too!

What's next for FlareTracker

There are still some things I want to finish and polish up - the AI integration needs to be proxied through a Supabase edge function and I need to use proper storage for the images.

I'd really like to launch FlareTracker in the very near future.

Built With

• bolt.new
• chatgpt
• copilot
• elevenlabs
• javascript
• netlify
• supabase