Inspiration
During a family visit, I watched my grandmother repeatedly ask my aunt the same question about her children five times in one hour. Each time, my aunt patiently answered with the same warmth and detail, but I could see the emotional toll it was taking on her. That moment made me realize: what if AI could provide that same patient, compassionate support 24/7? With 55 million people worldwide living with dementia and their families facing burnout, I knew AWS Bedrock's advanced language models could help. Memory Companion was born from a desire to maintain dignity for patients while giving caregivers the support they desperately need.
What it does
Memory Companion is an autonomous AI care assistant specifically designed for dementia patients. It provides: For Patients: Natural, empathetic conversation that never gets frustrated with repeated questions For Caregivers: Real-time dashboard with conversation insights
How we built it
How we built it AWS Architecture: Amazon Bedrock with Claude 3.5 Sonnet Development Stack: Frontend: React with Tailwind CSS Backend: Node.js with Express for API handling AI Integration: AWS SDK for Bedrock with Claude 3.5 Sonnet Database: In-memory storage (designed for DynamoDB migration)
Challenges we ran into
Bedrock Integration Complexity
Accomplishments that we're proud of
The system doesn't just answer questions; it understands the emotional context of dementia care and responds with genuine warmth.
What we learned
Building with empathy as the core principle, not just functionality
What's next for Memory Companion
Photo recognition so patients can take pictures of family members and get instant context
Log in or sign up for Devpost to join the conversation.