Alzyra

• 

Inspiration

The inspiration behind Alzyra came from a deeply personal place — my grandmother has been struggling with memory loss and often forgets daily events or familiar faces. Watching her confusion and frustration made us realize how isolating and painful this experience can be, not just for patients but also for their families.

Our team wanted to create something with real social impact, something that could help people like her hold onto their memories, stay independent longer, and bring peace of mind to caregivers. That’s how Alzyra – an AI-powered memory reconstruction companion – was born.

What It Does

Alzyra helps people with early-to-moderate Alzheimer’s and dementia actively reconstruct memories instead of just being reminded of them.

Through AI-guided conversations, the app gently asks guided questions — like “Do you remember if you were indoors or outdoors?” — and uses contextual cues from the user’s phone (photos, location, and calendar events) to help them recall experiences naturally.

It also features a caregiver portal that provides insights into memory patterns, medication adherence, and emotional trends. In short, Alzyra is a compassionate digital companion that helps preserve identity, strengthen memory, and support caregivers.

How We Built It

We combined advanced AI with human-centered design to make Alzyra both smart and empathetic.

Core technologies used:

LiveKit + Vapi AI for real-time, natural voice interactions

Claude / OpenAI for contextual understanding and emotional tone adaptation

Device integrations with photos, GPS, and calendar for memory cues

Dual interface system:

Patient mode — large touch targets, minimal cognitive load

Caregiver mode — analytics dashboard with memory strength and behavioral insights

Privacy-first architecture ensures all personal data is encrypted and consent-driven.

Challenges We Ran Into

Designing for cognitive decline while keeping the experience empowering, not infantilizing

Balancing empathy and accuracy — making the AI sound compassionate, not robotic

Managing privacy for sensitive memory and health data

Translating neuroscience principles like spaced repetition into AI dialogue patterns

Voice response latency and adaptation for older users with slower reaction times

Accomplishments That We’re Proud Of

Built a functional prototype capable of guiding real-time recall conversations

Developed a context-aware AI model that adapts to user memory patterns

Created a dual-portal system linking patient interaction with caregiver insights

Received positive feedback from early testers and caregivers

Stayed true to our mission — technology with social and emotional purpose

What We Learned

Effective memory care technology must empower recall, not automate it

Designing for dementia requires empathy, patience, and iteration

Cross-disciplinary collaboration between AI, neuroscience, and UX is crucial

Emotional tone is as vital as accuracy — trust builds engagement

Compassion-driven technology can truly enhance quality of life

What’s Next for Alzyra

Launch pilot studies with Alzheimer’s and dementia care clinics

Expand memory reconstruction models to support multiple languages and cultures

Integrate wearables and biometrics for advanced cognitive tracking

Partner with healthcare organizations and insurers to scale accessibility

Continue developing Alzyra into the world’s most trusted AI companion for memory health

Built With

• claude
• expo.io
• javascript
• livekit
• openai
• react-native
• vapi