MediSage AI

• 

  asking a question

• 

  the answer we got

• 

  landing page

Inspiration

The idea came from my mom. She was searching on the internet about a health problem I had. She visited many random websites. The information was confusing and not always true. I realized many people do this every day. I wanted to make something better. Something that gives correct and clear answers. So I built MediSage AI.

What it does

MediSage AI helps people understand medical images and health questions. You upload a photo of a medical image. You can also ask a health-related question. The AI looks at your image and your question. It gives a smart and helpful answer. It is useful for medical students, doctors, and curious people.

How we built it

We used the Perplexity API to answer questions. We used image input to let users upload medical images. We built the app using TypeScript and React. We designed it to be fast and easy to use. We tested it with different images and questions.

Challenges we ran into

Medical topics are hard. We had to be careful with how we answer. We had to test many images to make it work well. Making the interface simple but powerful was also hard. Sometimes the AI needed help understanding unclear questions.

Accomplishments that we're proud of

We made a working demo in a short time. It can analyze images and give useful answers. It feels helpful and friendly. We turned a personal problem into a real solution. We used AI in a smart and meaningful way.

What we learned

We learned how to use Perplexity's API for real problems. We learned how to work with image input and question answering. We learned how hard medical topics can be. We also learned how much people want clear and trusted answers.

What's next for MediSage AI

We want to improve the accuracy. We want to add more types of medical data. We want to add safety checks and warnings. We may partner with doctors and health platforms.

Built With

• sonar
• typescript
• vite