The official website of Rafa and Jess Lyóvson
A personal family blog combined with a public zettelkasten (digital knowledge garden) where we share our thoughts, ideas, and intellectual journey as a family.
This site serves as both:
- Family Blog: Articles, reviews, and project updates from Rafa and Jess
- Public Zettelkasten: A connected web of notes, ideas, and thoughts that evolve over time
- Knowledge Base: References to books, movies, and other media with our family's thoughts and quotes
- Dynasty Archive: A growing collection of family wisdom that will expand as our family grows
- Frontend: Next.js 15+ (App Router) with React 19+ Server Components
- CMS: Payload CMS v3 (Headless CMS)
- Database: Neon Postgres with pgvector for semantic search
- Styling: Tailwind CSS + Shadcn UI components
- Animations: Motion for React
- Rich Text: Lexical Editor
- Email: Resend
- Hosting: Vercel
- Vector Embeddings: pgvector 0.8.0 (VARCHAR storage with runtime casts)
- Semantic Search: Powered by text-embedding models
- Content Discovery: AI-driven connections between posts and notes
Articles, reviews, and project updates written by family members
- Types: Articles, Reviews, Project Updates
- Features: Rich text content, semantic search, cross-references
Zettelkasten-style notes with different types and purposes
- Permanent Notes: Developed concepts and ideas
- Literature Notes: Quotes and thoughts from books/media
- Fleeting Notes: Quick captures and temporary thoughts
- Index Notes: Navigation and organization
Curated collections of external content with API integrations and family curation
- Books: Google Books API integration with authors, genres, publication data
- Movies: OMDB API integration with directors, actors, cast information
- TV Shows: Episode tracking, creators, cast, and seasonal organization
- Music: Albums and songs with artists, genres, and streaming platform links
- Podcasts: Episodes with hosts, guests, and platform availability
- Video Games: Gaming collection with developers, platforms, and genres
- People: Central registry of creators, authors, directors, and other contributors
- Links: Curated web content, articles, and casual video references
The People collection serves as a central registry for all creators and contributors:
- Multiple Roles: Authors, directors, actors, hosts, musicians, developers
- Cross-Media Relationships: One person can be linked to books, movies, podcasts, etc.
- Bidirectional Connections: People → Media and Media → People relationships
- Literature Notes: Can reference people directly for quotes and thoughts
- Reading/Watching/Listening Status: Track progress across all media types
- Family Ratings: Shared scores and opinions on content
- Quote Collections: Meaningful passages with family commentary and analysis
- Cross-References: Connect media to posts, notes, and other references
- Related Content: AI-powered suggestions while writing
- Cross-Author Insights: Find where family members wrote about similar themes
- Unexpected Connections: Discover links between seemingly unrelated ideas
- Literature Integration: Connect quotes to original thoughts
- Multi-Author System: Content from both Rafa and Jess
- Family Ratings: Shared opinions on books, movies, and shows
- Cross-References: Manual and automatic connections between content
- Knowledge Evolution: Track how ideas develop over time
- Flexible Relationships: Notes can connect to posts, other notes, and references
- Status Tracking: Reading/watching progress for media
- Automatic Metadata: Generated embeddings, timestamps, and hashes
- Version Control: Draft and published states with version history
- Server-Side Rendering: Fast initial page loads
- Minimal JavaScript: Progressive enhancement for interactivity
- Responsive Design: Mobile-first with elegant desktop experience
- Accessibility: WCAG compliance and semantic HTML
- Atomic Notes: Each note represents a single idea
- Linking: Extensive cross-referencing between concepts
- Emergence: Ideas develop through connections over time
- No Hierarchy: Organic organization rather than rigid structure
# Install dependencies
pnpm install
# Set up environment variables
cp .env.example .env.local
# Start development server
pnpm dev- Configure Neon Postgres connection
- Run Payload CMS migrations
- Set up pgvector extension
- Embeddings auto-generate on publish (no manual setup needed)
- Access Payload CMS admin at
/admin - Create posts and notes through the CMS
- Embeddings are generated automatically
- Semantic search works immediately
- Content: Optimized for thousands of posts and notes
- Performance: Fast semantic search queries (~2.7ms at current scale)
- Storage: Efficient vector storage with Neon Postgres
- Costs: Minimal overhead with serverless architecture
- Multi-generational: Designed to grow with the family
- Knowledge Accumulation: Years of connected thoughts and ideas
- Cross-generational Discovery: Find connections across time periods
- Archive Stability: Built for long-term preservation
# Database
DATABASE_URL=postgresql://...
# Payload CMS
PAYLOAD_SECRET=...
# OpenAI (for embeddings)
OPENAI_API_KEY=...
# Email
RESEND_API_KEY=...posts- Family blog articles and reviewsnotes- Zettelkasten notes with types and literature notesbooks- Book references with Google Books API integrationmovies- Movie references with OMDB API integrationtvShows- TV show references with episode and season trackingmusic- Albums and songs with streaming platform integrationpodcasts- Podcast episodes with hosts and platform linksvideoGames- Gaming collection with platforms and developerspeople- Multi-role creator registry (authors, directors, artists, etc.)links- Curated web content and casual references
- Create content in Payload CMS
- AI suggests related existing content
- Add manual connections between ideas
- Publish to generate embeddings
- Discover unexpected connections over time
- Literature Notes: Save meaningful quotes with commentary
- Idea Development: Grow fleeting notes into permanent concepts
- Cross-Pollination: Find connections between family members' thoughts
- Wisdom Accumulation: Build a searchable family knowledge base
- Architecture Details: See
/docs/architecture.md - Embedding Strategy: See
/docs/new-schema-2.md - Design System: See
/docs/design-system.md - Implementation Guide: See
/docs/implementation-guide.md
This site is designed to be a multi-generational family knowledge base - a place where ideas, thoughts, and wisdom accumulate over time. As our family grows, so will the connections between concepts, creating an ever-richer web of family intellectual history.
The combination of modern web technology with timeless zettelkasten principles creates a platform for long-term knowledge building that will serve the Lyóvson family for decades to come.
Built with ❤️ by the Lyóvson family