How malikyakub/alignerai works
Project type and purpose
• This is a single-page web application (SPA) built with React (Vite) and Tailwind CSS.
• It uses Appwrite as a BaaS for authentication and database storage.
• It embeds simple AI logic (recommendations, alignment, time conversion) as JavaScript modules in the public/Models folder.
• It will be hosted as static files (e.g., Vercel) with client-side routing.
Main components
a. Browser / Client (React SPA)
b. In-Browser AI Engine (public/Models: Aligner.js, recommendator.js, time_converter.js)
c. Appwrite Backend (Auth, Database)
d. Static File Hosting (Vercel or any CDN)
e. External Services (if any—in this case, Appwrite is the main external)
Relationships & data flows
• User interacts with the React UI (src/components + src/pages)
• UI calls Appwrite SDK (src/appwrite/config.js, database.js) to CRUD tasks
• Data is stored in and retrieved from Appwrite Database
• After retrieval, UI feeds task list to the AI Engine modules (direct function calls)
• AI Engine returns the “next task” recommendation back to the UI
• UI presents results to the user
Architectural patterns & design principles
• SPA architecture with clear separation:
• Presentation layer (React components, pages)
• Service layer (Appwrite SDK wrappers in src/appwrite)
• Model layer (AI modules in public/Models)
• Component-based UI (Atoms/Molecules/Organisms in src/components)
• Client-side data flow: REST/RPC calls to Appwrite, then local processing