How webdevsimplified/ai-powered-job-prep works
Project type and purpose
• This is a full-stack, server-rendered Next.js application with both UI code and API routes living under src/app.
• The README confirms it was scaffolded via create-next-app and runs as a Next.js app on Vercel (or Docker Compose).
• Business domain: a job-interview management portal. Users sign in (via Clerk), create “jobInfos”, schedule interviews, generate AI-powered questions and feedback, and analyze resumes using external AI/Hume services.
File structure
• Top-level:
• app/ — UI pages and Next.js App Router.
• app/api/ — Next.js Route Handlers for serverless APIs (AI endpoints, webhooks).
• components/, features/ — React UI components and feature modules (actions, db caches, permissions).
• services/ — Thin wrappers around external services: Clerk (auth), AI providers (Google, Hume), in-house AI orchestration.
• drizzle/ — Drizzle ORM config, schema definitions, migrations for the relational database.
• docker-compose.yml & drizzle.config.ts — local dev and DB setup.
• Patterns:
• Layered architecture: Presentation (Next.js pages + components) → Application logic (feature modules + services) → Data access (Drizzle ORM) → External systems (Auth, AI, DB).
• File-based routing → API endpoints under app/api/….
README for additional insights
• Deploy instructions hint at Vercel but docker-compose.yml implies local containerization (Postgres, possibly Redis).