How jitenverma076/talent-arc works
Project type and purpose
• TalentArc is a single-page web application (SPA) built with React and Vite that matches technical candidates to companies via skill-based challenges.
• It leverages Firebase for authentication and data storage (Firestore) and Google Gemini AI for candidate analysis.
File structure
• src/
• components/ – Reusable UI components (Navbar, Footer, AuthModal, ThemeToggle, etc.)
• pages/ – Route-level components (Home, Dashboard, Challenges, Companies, Profile, AdminDashboard, etc.)
• context/ – React Context providers (AuthContext, ModalContext)
• hooks/ – Custom hooks (useAuth)
• lib/ – Abstractions over Firebase (auth.js, firestore.js) and utility functions
• utils/ – Static or demo data (candidateData.js)
• assets/ – Images and static files for the UI
• main.jsx & App.jsx – Application entry point and root layout
• Public folder & index.html – Static host file
• Configuration files – Vite, Tailwind, ESLint, .env.example
• firestore.rules – Security rules for Firestore
Key architectural insights from README
• Tech stack: React 19 (Vite + Tailwind CSS), Firebase Auth + Firestore, Google Gemini AI (fallback to statistical analysis)
• Authentication & API integration points: Firebase SDK, HTTP calls to Gemini AI
• No separate “backend” service other than Firebase and external AI
Main components
a. Client SPA (React/Vite)
• UI Layer: Pages, Components, Context Providers, Hooks
• Routing & Protected Routes
b. Firebase services
• Authentication (Email/Password)
• Firestore (Challenges, Users, Results)
• Firestore Security Rules
c. External AI Service
• Google Gemini AI API (via REST/HTTPS)
• Fallback statistical analysis module (in src/lib/utils.js)
d. Build & Deployment
• Vite Dev Server & Production Build
• Hosting (e.g., Firebase Hosting or any static host)
Relationships & interactions
• The SPA renders pages and components, uses AuthContext to manage user sessions.
• On login/register flows, it calls Firebase Auth SDK.
• Authenticated sessions fetch user and challenge data from Firestore.
• Submissions or candidate profiles trigger calls to Gemini AI API.
• AI responses are stored/displayed via Firestore or in-memory state.
• UI responds to data via React state and Context.
Architectural patterns & design principles
• SPA (client-heavy) with thin “backend” (Firebase) and pluggable AI service
• Separation of concerns:
• UI components (pure presentation)
• Context for global state (Auth, Modal)
• Hooks for logic reuse (useAuth)
• lib/ for service abstraction (firebase.js, firestore.js, auth.js)
• Protected routes & error boundaries for resilience
• Configurable via .env for different environments