Inspiration
We were inspired to develop this project through our members’ experiences supporting their elderly family members. Especially with those at long distances, supporting their medical needs can be a difficult task. Furthermore, we learned from literature review that 69% of seniors reported making mistakes when taking their medication in the past 6 months, and 73% of seniors cited memory struggles as the reason for noncompliance to medicine regimens (Ghassab-Abdollahi et al.; Rodgers et al.). This inspired us to develop a solution targeting this issue, to support seniors in safely taking their medications.
What it does
MedAssist keeps a log of user medications and helps to make sure elderly people take their medicines at the appropriate time with the help of their caregiver. Users have the option to name, upload, or take an image of the respective medication they wish to add. When scanned, MedAssist identifies the medication information rapidly. MedAssist automatically restricts dangerous and unreasonable submissions. If the submission is flagged by the system as potentially dangerous, an approval request is sent to their respective caretaker to verify that the provided medical regimen is, indeed, authorized. The caretaker can then choose to authorize or deny the request, with immediate notification. The caretaker can further update approved medications for their respective patient.
How we built it
We utilized Gemini and Perplexity to perform research on competitor products and current solutions. Claude was used for developing and formatting our Project Requirements Document (PRD). To design our User Interface (UI), we used Lovable and Replit Agent 4, producing a total of 6 potential screen designs. We brainstormed app names with ChatGPT and created logo proposals using Ace Logo Maker, with a total of 6 potential logos. We selected the name, MedAssist, alongside our logo and UI.
Throughout the period of the hackathon, we used Replit and the Task Manager to collaboratively add and develop features on our app. Periodically, we synced our code to a GitHub repository.
Testing was performed to check the functionality of the following: account, medication scanning, safety checks, and caretaker authorization. Furthermore, tests were repeated for three different sets of users and caretakers to ensure proper app function.
Challenges we ran into
There was difficulty in establishing the patient-caretaker connection to ensure effective approval in a timely and secure manner. Furthermore, we faced the question of which caretaker verification is required for medications. Additionally, there were numerous feature-related issues that we addressed throughout the development process.
Accomplishments that we're proud of
We are proud of our caregiver and patient system that allows the elderly to stay safe by confirming their medicines and automatically informing their caregiver.
What we learned
We learned to use the task manager on Replit to collaborate while developing MedAssist.
What's next for MedAssist
Our next steps in developing MedAssist will involve ensuring the security of our app via a thorough scan utilizing SecureVibes, which uses Claude to identify the app. Furthermore, we will create an IOS app to allow for ease of installation and use by a broader audience.
Acknowledgments
We were trained on Vibe Coding by our mentors at the Aidya (https://theaidya.com/).
References
Ghassab-Abdollahi, Nafiseh, et al. "Self-administration medication errors at home and its predictors among illiterate and low-literate community-dwelling older adults with polypharmacy: a negative binomial hierarchical regression." PLoS One 19.4 (2024): e0302177. Rodgers, Jo E., et al. "Predictors of medication adherence in the elderly: the role of mental health." Medical Care Research and Review 75.6 (2018): 746-761.
Log in or sign up for Devpost to join the conversation.