Find Quotes
ActiveQuote discovery platform with emotion-based search, artist filtering, and custom image generation
Overview
Find Quotes is a quote discovery platform that helps users find the perfect quote through emotion tags, artist filtering, and exact text matching. It includes a custom image generator to create shareable quote graphics.
The Problem
The initial version used RAG (Retrieval-Augmented Generation) with in-memory embedding models for semantic search. While powerful, it consumed excessive resources since everything ran in-memory without external embedding services. Without real users, the infrastructure costs couldn't be justified.
The Solution
Rebuilt as a lean Next.js application deployed on Vercel:
- File-based Storage: Quotes stored in static files for zero database overhead
- Multiple Search Methods: Emotion tags, artist filtering, and exact text matching
- Image Creation: Custom quote graphics with 6 template options
- Serverless Architecture: Pay-per-use model that scales with actual usage
Key Features
Search & Discovery
- Emotion Tags: Find quotes by mood (inspirational, melancholic, motivational, etc.)
- Artist Filter: Browse quotes from specific authors
- Text Search: Exact string matching for finding specific phrases
Image Generator
- 6 Template Options: Choose from different visual styles
- Quote Placement: Position text on images for optimal readability
- Download: Export custom quote images for sharing
Technical Highlights
- Frontend: Next.js for fast, SEO-friendly rendering
- Deployment: Vercel for serverless hosting
- Design: Lightweight, cost-effective architecture
Current State
The app is live and functional at find-quotes.com with a curated collection of example quotes. The focus is on getting users first before investing in advanced features.
Future Plans
- →Get actual users first
- →Add RAG-based semantic search when user base justifies resources
- →Implement AI image generation to match quotes
- →Expand quote database significantly
Update Log
Added seo for all quotes and resubmit sitemap to google and bing
Rebuilt as lightweight Next.js app with file-based quotes and image creation features
Initial version with RAG and semantic search (archived due to resource usage)