This document outlines the strategic vision for the next phases of development for the Recipe Generator platform.
- LLM Integration: Transition from static Transformer models to dynamic LLMs (like GPT-4o-mini or Llama 3) for more creative and natural-sounding cooking instructions.
- Voice Assistant: Implement a hands-free "Cooking Mode" using Web Speech API to read instructions aloud.
- Dietary Profiling: Advanced filtering that automatically excludes allergens (nuts, gluten) from generated instructions based on user profiles.
- Public Recipe Feed: A "Discover" page where users can share their high-quality generated recipes.
- Rating System: Allow the community to rate and comment on the accuracy of AI-generated recipes.
- Chef Profiles: Public user profiles displaying a curated collection of "Signature Dishes."
- PWA Support: Transform the site into a Progressive Web App for offline access to the shopping list.
- Pantry Manager: A feature to track what ingredients you already have, allowing the AI to suggest recipes based on "Image + Inventory."
- Kitchen Timer Integration: Build-in timers directly into the recipe steps for easier multi-tasking.
- Managed Database: Migration from SQLite to PostgreSQL (Supabase or AWS RDS) for better performance and scaling.
- Image Lifecycle Management: Automated cron jobs to clean up
demo_imgsolder than 24 hours to save server space. - Global CDN: Deploy static assets via Cloudflare to ensure fast load times for international users.
If you're a developer interested in picking up one of these "Future" tasks, please check our GitHub Issues or reach out via the community discussion board!