Inspiration

We were inspired by the growing need for continuous support for dementia patients, who often feel isolated and struggle with memory loss. We realized that caregivers can get overwhelmed and could benefit from automation to focus on more critical needs.

What it does

The program includes a caregiver portal where patient information can be input and updated. When the patient verbally asks a question, their speech is converted to text, and using Gemini-2.0-Flash along with data from the portal, the AI generates a verbal response (text-to-speech). The patient’s mood and any new information they share are analyzed, and the caregiver portal is updated accordingly.

How we built it

Created a dual-threaded Python application that runs a Flask web server for caregivers to monitor and update patient data. It handles real-time patient interaction via speech-to-text, AI processing (Gemini-2.0-Flash), and text-to-speech.

Challenges we ran into

Accomplishments that we're proud of

Developed an efficient speech recognition system that transcribes patient speech and leveraged Gemini-2.0-Flash to generate tailored responses that correctly addressed the patient’s needs and history.

What we learned

What's next for AI Dementia Assistant

Built With

Share this project:

Updates