Fragments of Me
An immersive digital reading and writing sanctuary featuring ambient soundscapes and line-by-line reactions.
Overview
Fragments of Me is an emotion-driven reading and writing environment designed to evoke contemplation and personal expression. Rather than treating writing as content for consumption, the platform prioritizes atmosphere through mood-based soundscapes, granular line-by-line reader reactions, and an editorial author ecosystem.
Architecture
Storefront & Reading UI
Dedicated Microservices
Data & Auth (Firebase)
Engineered Features
Ambient Soundscapes
Integrated web audio engine offering 4 curated mood environments (warm, soft, melancholic, dark) with seamless looping and persistent volume controls across route transitions.
Line-by-Line Reactions
Readers can leave contextual emoji reactions on specific lines. Each line maintains its own isolated subcollection in Firestore to prevent hitting document size limits.
Echoes Commentary
Threaded reflective discussions allowing readers to exchange insights and interpretations on essays and poems.
Daily Ritual Tracking
Consistent reading and writing streak counter encouraging continuous creative reflection without intrusive notifications.
Technical Implementation
- Built with Next.js 16 and React 19, taking advantage of improved hydration speed and modern hooks.
- Utilizes Firestore Subcollections to store fine-grained reader reaction counts, ensuring document reads remain O(1) and never exceed the 1MB Firestore document limit.
- Separated heavy email dispatch tasks into an isolated Express.js microservice to prevent long-running SMTP handshakes from blocking client web workers.
- Implemented strict Content Security Policies, security response headers, and dynamic SEO sitemaps.
Lessons Learned
Designing for emotional engagement requires equal parts aesthetic sensitivity and data design. Designing the line-by-line reaction feature demonstrated the importance of subcollection partitioning in NoSQL databases, maintaining instant UI optimistic updates even when readers react in quick succession.