PillSync

Your personalized birth control companion Track your cycle, check medication interactions, and stay on top of your birth control routine with confidence.

Comment

Inspiration

This project was inspired by one of our team members, Advitiya. She has PCOS, an uncurable condition that impacts millions of women. Like most people with PCOS, Advitiya was recommended to go on birth control. However, soon after starting the pill, she experienced pain and discomfort due to the reaction of the pill and the iron supplements she was taking. Advitiya's experience is an example of how common medication reactions can be for women with PCOS. This inspired the idea of PillSync.

What it does

PillSync is a user friendly app for all women on the pill and is not exclusive to those with PCOS. Its features include daily reminders to take the pill, a personalized flow calendar, and visual representations of the user's recent ovulation suppression, performance and energy levels. Additionally, it has a medication interaction checker, allowing users to input any medication or supplement and check if it has an effect on their birth control. The goal of PillSync is to act as a reliable birth control companion and help women stay consistent, informed, and confident in their routine.

How we built it

We used React to develop both the front end and back end, incorporating many UI primitives such as the Accordion, Alert Dialog and Aspect Ratio components to ensure a user friendly and interactive interface. Our backend includes an embedded AI layer that handles the probability calculations for the medication interaction checker. Additionally, we used Figma AI to increase efficiency. Three AI platforms were used:

  1. RxNav AP1: A medical database.

  2. Embedded AI: Connects symptoms to mathematical vectors and tries to match them their respective medication reactions using database.

  3. Open AI: Summarizes the information after confirming its accuracy.

Challenges we ran into

We ran into a series of challenges including:

  1. Cases where the AI would hallucinate
  2. Trouble storing user data and integrating user data into final analysis
  3. Manual entry to change languages on front end coding
  4. Creating a user friendly UI for a wide range of ages (simple and informational)
  5. Simplifying or breaking down medical terminology

Accomplishments that we're proud of

  1. Used three AI platforms to overcome challenges faced
  2. Used mathematical vectors to calculate accuracy of reactions before sending to Open AI

What we learned

We learned how to eliminate AI hallucinations by incorporating more one AI platform

What's next for PillSync

  1. Link it to other user accounts (family members)
  2. A medical history upload option

Built With

Share this project:

Updates