Skip to content

ARCHITECTURE.md

Latest commit

 

History

History
90 lines (73 loc) · 3.88 KB

ARCHITECTURE.md

File metadata and controls

90 lines (73 loc) · 3.88 KB

CareerCraft Architecture

System Overview

CareerCraft uses a Service-Oriented Architecture (SOA) with a unified frontend orchestrating interactions between a persistent Backend API and a stateless ML Service. This hybrid approach enables high-performance AI inference without overloading the main application server.

Key Components

  1. Frontend (Next.js) (Port 3000)

    • Role: Main user interface and orchestrator.
    • Tech Stack: Next.js 16, React 19, Tailwind CSS 4, Lucide React.
    • Responsibilities:
      • Rendering UI (Server & Client Components).
      • Managing Authentication state (Context/Cookies).
      • Communicating with Backend for user data.
      • Communicating directly with ML Service for real-time analysis.

  2. Backend (Node.js/Express) (Port 5000)

    • Role: Persistent data manager and authentication provider.
    • Tech Stack: Express.js, Mongoose, JWT, Bcrypt.
    • Database: MongoDB (User data, Profiles, Saved Cover Letters).
    • Storage: Cloudinary (Resume file storage).
    • Responsibilities:
      • User Authentication (Signup/Signin).
      • Profile Management.
      • storing and retrieving saved resources (Resumes, Cover Letters).

  3. ML Service (FastAPI) (Port 8001)

    • Role: Stateless intelligence engine.
    • Tech Stack: Python, FastAPI, Spacy, Scikit-Learn, Sentence-Transformers.
    • Responsibilities:
      • Resume Parsing (Text Extraction from PDF/DOCX).
      • Resume Analysis (Keyword scoring, feedback).
      • Job Matching (Cosine similarity).
      • Cover Letter Generation (Template/GenAI logic).

High-Level Architecture Diagram

+----------------------+       +----------------------+
|   User Client        |-------|   Third-Party Services|
|  (Browser/Next.js)   |       | (Cloudinary/MongoDB) |
+----------------------+       +----------------------+
       |          |
       | (Auth /  | (Analysis /
       | Persistence) Inference)
       v          v
+------------+  +-------------+
|  Backend   |  | ML Service  |
| (Express)  |  |  (FastAPI)  |
+------------+  +-------------+
       |
+------------+
|  Database  |
| (MongoDB)  |
+------------+

Data Flows

1. Resume Analysis

  • User Action: Uploads a resume file on "Resume Analysis" page.
  • Flow: Frontend -> ML Service (POST /resume/extract-text & /resume/analyze).
  • Result: Instant feedback displayed to user. No persistent storage in DB until explicitly saved.
  • Rationale: Reduces load on Backend; keeps analysis fleeting and fast.

2. Resume Upload (Profile)

  • User Action: Uploads a resume on "Upload Resume" page.
  • Flow: Frontend -> Backend (POST /api/resumes/upload).
  • Result: File stored in Cloudinary, link saved in MongoDB User Profile.

3. Cover Letter Generation

  • User Action: Enters job details and analyzes resume.
  • Flow: Frontend -> ML Service (POST /cover-letter/generate-cover-letter).
  • Result: Generated text returned to Frontend.
  • Save Action: User clicks "Save" -> Frontend -> Backend (POST /api/cover-letters).

Security & Authentication

  • JWT Strategies: Backend issues JWTs upon login. Frontend includes Bearer tokens in requests to Backend.
  • ML Service Security: Currently open (CORS allowing all for development) or protected via internal tokens (Future improvement).
  • Environmental Variables: All sensitive keys (DB URI, Cloudinary Secrets) are stored in .env files and never exposed to the client.

Scalability Considerations

  • Stateless ML: The ML Service is stateless, making it easy to scale horizontally on GPU/CPU-optimized instances.
  • CDN: frontend assets served via Next.js optimization.
  • Database: MongoDB handles flexible schema for evolving user profiles.