Find your path. Build your future.
PathPilot AI is an interactive career-guidance platform built for students, freshers, and early-career professionals. It turns the overwhelming question of “what should I learn next?” into clear career exploration, structured learning roadmaps, side-by-side comparisons, and contextual AI mentoring.
Career information is usually fragmented: role descriptions live in one place, learning resources in another, and personalised guidance is difficult to access. This makes it hard for learners to confidently choose a direction and make consistent progress.
PathPilot AI brings the decision-making journey into one focused experience. Learners can explore curated career paths, understand the skills each path needs, follow a roadmap, compare options against their priorities, save promising careers, and ask an AI mentor for guidance.
- Career discovery — browse 20+ curated career paths, filter by category and difficulty, or search by role, skill, and category.
- Career details and roadmaps — view role overviews, skills, learning duration, salary ranges, demand indicators, projects, resources, and step-by-step learning paths.
- Skill explorer — learn what each skill is, why it matters, and where it fits in a career path.
- Career comparison — compare two roles with visual metrics such as salary, learning time, job demand, and difficulty; request an AI comparison for additional context.
- AI career mentor — ask natural-language questions about careers, skills, and learning plans through a Gemini-powered assistant.
- Progress and motivation — track roadmap progress, maintain a learning streak, receive daily tips, and celebrate milestones.
- Personal shortlist — save favourite careers locally and revisit them whenever needed.
- Polished, responsive UX — responsive layouts, animated transitions, accessible controls, and a dark glassmorphism interface.
| Area | Technologies |
|---|---|
| Frontend | React 18, Vite, React Router |
| Styling | Tailwind CSS, PostCSS |
| Motion & visuals | Framer Motion, Lucide React, Canvas Confetti |
| Backend | Node.js, Express, CORS |
| AI | Google Gemini API (@google/generative-ai) |
| Persistence | Browser localStorage for favourites, roadmap progress, and streaks |
React + Vite client (port 5173)
|
| /api requests (Vite proxy in development)
v
Express API (port 8787)
|
| Gemini API key stays on the server
v
Google Gemini
Career data, roadmaps, learning resources, daily tips, and comparison metrics are curated local data files, so the core exploration experience remains available without an AI request. The backend accepts the AI requests and keeps Gemini credentials out of the browser bundle.
- Node.js 20 or later
- npm
- A Google Gemini API key (only required for AI features)
Install the frontend dependencies:
npm installInstall the backend dependencies:
cd server
npm install
cd ..Create server/.env and add your Gemini API key:
GEMINI_API_KEY=your_gemini_api_key
# Optional: use a different supported Gemini model
GEMINI_MODEL=gemini-2.5-flash
# Optional: comma-separated frontend origins for deployed clients
FRONTEND_URL=http://localhost:5173To use a separately deployed backend, copy .env.example to .env in the project root and set its public URL:
VITE_API_URL=https://your-api.example.comLeave VITE_API_URL empty for local development; Vite proxies /api calls to http://localhost:8787.
Never commit real API keys. Root
.envfiles are ignored by Git; keepserver/.envprivate as well.
Start the API in one terminal:
cd server
npm run devStart the frontend in another terminal:
npm run devOpen the URL Vite prints (normally http://localhost:5173). The API health check is available at http://localhost:8787/api/health.
| Location | Command | Purpose |
|---|---|---|
| Project root | npm run dev |
Start the Vite development server |
| Project root | npm run build |
Create a production frontend build in dist/ |
| Project root | npm run preview |
Preview the production frontend build |
server/ |
npm run dev |
Start the API with file watching |
server/ |
npm start |
Start the API normally |
Deploy the frontend as a static Vite site and the server/ directory as a Node/Express service.
- Deploy the backend and set
GEMINI_API_KEY,GEMINI_MODEL(optional), andFRONTEND_URLto the URL of the deployed frontend. - Set
VITE_API_URLin the frontend deployment environment to the backend’s public URL. - Build the frontend with
npm run buildand deploy the generateddist/directory.
The server also supports numbered Gemini keys such as GEMINI_API_KEY_1 and GEMINI_API_KEY_2; it attempts available keys in order, which can help with quota resilience.
PathPilotAI/
├── src/
│ ├── components/ # UI, layout, charts, roadmap, and AI components
│ ├── data/ # Curated careers, skills, roadmaps, resources, and tips
│ ├── hooks/ # Progress, streak, and favourites hooks
│ ├── pages/ # Route-level screens
│ └── utils/ # Data access, AI client, and local-storage helpers
├── server/
│ └── index.js # Express API and Gemini integration
├── public/ # Static assets
└── vite.config.js # Vite configuration and local API proxy
PathPilot AI is designed to make career planning more actionable and less intimidating. Instead of presenting learners with generic advice, it combines curated path data with interactive progress tracking and an on-demand AI mentor. The result is a practical starting point for making informed, personalised career decisions.
- User accounts and cloud-synced progress
- Personalised recommendations based on goals and skill assessments
- Verified course, certification, internship, and job links
- Mentor communities and feedback loops
- Analytics that adapt roadmaps to learner progress
This project is currently unlicensed. Add a license before using or distributing it beyond the hackathon context.