Memento Assistant

SYNOPSIS
Memento is a virtual memory app designed to assist caregivers and Alzheimer’s patients by preserving cherished memories in an accessible, user-friendly format.

As the lead for UX/UI and front-end development, I collaborated with a team to build the app in 24 hours during the Revolution UC Hackathon.

Integrating a custom AI backend, we aimed to provide a meaningful tool that aims to enhance memory retention and emotional connection. With our AI model, we were able to create an AI with a "long term memory".

Our project won 3rd place overall and received multiple awards, including Best Social Impact Hack and sponsor recognition.
TYPE
Hackathon
Front End Ideation
Figma Prototype
UX/UI
CREDITS
Arya Garg
Raihan Rafeek
Kaaustaaub Shankar
DATE
February 2024

DIALECTIVE

DIALECTIVE is a New York City-based fashion collective that empowers emerging independent designers through community-driven support, creative showcases, and exclusive opportunities like NYFW showrooms.

The Inspiration

The inspiration for Memento stemmed from a deeply personal experience. In 2023, my grandfather Jere was diagnosed with Alzheimer’s disease, leading my grandmother Judy to become an active advocate in Cincinnati.

Through conversations with my grandmother, I realized the many unmet needs for those living with Alzheimer’s, particularly around fostering independence and supporting caregivers.

Memento was designed with the goal of addressing these challenges, creating a tool that not only aids patients and caregivers but also contributes valuable data to Alzheimer’s research.

24 Hour Hackathon Goal

How might we design an AI powered journaling app that empowers individuals with Alzheimer’s and related dementias?

Information Architecture

Prototype Development

Onboarding & Journaling

The user is greeted with a brief introduction, a space to make custom journal entry templates, and is informed that they can opt in for their quantitative data to be shared with healthcare providers or caregivers.

Inserting Mementos

Users can add images, locations, voice recordings, and other media to their Mementos, which return back to their Memento Library.  

Memento Library

Once multiple Mementos have been added, the Memento Library organizes them into a visual journal that can be easily explored and discussed in the Chat tab.

Our custom AI analyzes these entries, identifying patterns and significant moments, while being able to recall specific data no matter how long ago it was added.

Quantitative data, such as instances where key information is repeatedly forgotten, and biometric data from smart devices can be flagged and automatically forwarded to caregivers (with user consent), ensuring timely intervention and better care management.

Make Some Mementos

Custom Long Term Memory AI 

After finalizing the Memento Figma mockup, I handed off the front-end design to the development team, enabling them to create a functional prototype.

This backend version integrates AI to analyze journal entries, recognizing patterns and providing insights to help both users, caregivers, and healthcare providers manage day-to-day activities. For further details, please visit the project onDevpost

Main Takeaways

Challenges Faced 🛠️
This project posed significant challenges, from creating an intuitive user experience to efficiently managing journal data. We initially used React Native through Expo but encountered issues with styling and iOS optimization. After a brief pivot to XCode and SwiftUI, we returned to React.js to build a mobile-scaled web app, using Figma to replicate the UX/UI.
Accomplishments 🎉
We successfully implemented core functionalities, including non-robotic, emotion-based responses that politely correct users when they are confused—something GPT-3.5 lacked at the time.
We showcased journaling, picture storage, and assistant querying with a 7-day dataset, and a significant achievement was vectorizing data tables in Tembo to efficiently return keyword-related entries.
Learnings 🧠
We learned a lot about app design on iOS (Swift), and front-end development, particularly how to design for usability across age groups.

Additionally, this was my first experience using Figma in a cross-disciplinary environment, where I learned to rapidly wireframe under tight deadlines, balancing the need for speed with design precision.
Next Project