Memoir is a digital memory preservation and personalized card creation app built to protect the moments that matter most. In a world where meaningful messages get lost in texts and screenshots, Memoir creates a dedicated space to scan, organize, and revisit handwritten cards and letters while also allowing users to design new personalized digital cards with music, animations, and voice notes.
By blending nostalgia with modern AI tools, Memoir preserves the past while helping create new moments worth saving.
- Scan and upload handwritten cards
- Text extraction using OCR
- Side-by-side display of scanned image + transcribed text
- Manual text correction
- Custom tagging
- Automatic AI-powered tagging
- Sorting by occasion
- Custom folder names
- Organized memory library
- Choose templates
- Customize fonts, colors, stickers
- Drag-and-drop design features
- Add music or voice notes
- Save and send digital cards
- Search by mood, occasion, tags, sender, keywords
- Vector search across OCR text + user tags
- Returns both individual cards and folders
- Timeline View → Display cards chronologically
- Reflections Section → Add personal notes
- Calendar + Reminders → Upcoming events + reminders
- Generative AI Card Designer → AI designs card from prompt
- Analytics Dashboard → Spotify Wrapped–style yearly recap
Week 1
- Finalize tech stack
- Assign roles
- Install software
- Frontend wireframing (Figma)
- Backend auth practice
Week 2
- Login / Signup pages
- Connect authentication
Week 3
- Search UI + card storage
- Database schema design
Week 4
- REST APIs
- Integrate Expo Camera + OCR
- UI polish
Week 5
- OpenAI tagging integration
- Presentation prep
Week 6
- Refine OCR
- Integrate vector search
Week 7
- Attempt stretch goals
Week 8
- Practice presentation
Week 9
- Final polish
Wireframing: Figma
IDE: VSCode
Version Control: Git
Frontend: React Native Expo (TypeScript)
- Tailwind via NativeWind
Backend: Supabase
- PostgreSQL
- Authentication
- Storage
- Edge Functions
AI:
- Google Cloud Vision API
- OpenAI API
- Supabase Vector Search
Core Stack
Styling + UI
Camera + OCR
AI + Search
Full Stack Setup
Authentication
Camera + Image Upload
Drag & Drop + Design
Backend + APIs
OCR Quality on Handwritten Cards
- Handwriting varies widely
- Solution: show OCR text beside image + allow manual edits
Automatic Tagging Accuracy
- AI may misinterpret emotion
- Suggest tags for approval instead of auto-assign
Drag and Drop Performance
- Too many animations may lag
- Use Gesture Handler + Reanimated
- Save card state as JSON
I read about elderly families who lost fifty years of handwritten letters and memories in the Palisades fire. It made me realize how fragile meaningful memories can be.
Almost everyone has a box filled with birthday cards and handwritten notes. They hold moments, not just words. But today, personal messages are reduced to texts and screenshots that disappear over time.
Memoir is meant to preserve those old memories while also encouraging people to continue creating thoughtful ones.
- Photomyne
- Canva
- Greetings Island
- Evite
- Paperless Post
Tamanna Khurana
Umaymah Sultana
- Jiya Khurana
- Tejasvi Annamaraju
- Harleen Gill
- Kasish Jain