AI-powered career guidance Android app — Gemini-driven career matching, MVVM architecture, Firebase backend.
Career Compass interviews the user about their skills, interests, strengths, and preferred work environment, then uses Google Gemini to generate ranked career-path recommendations — each with a fit percentage and a matching-factor breakdown — plus deep-dive analysis and small-business guidance for any chosen path.
- 🤖 Gemini-powered career matching — structured JSON responses parsed into ranked career cards with fit scores
- 💬 AI career chat — follow-up guidance conversation per career path
- 👤 Full auth flow — register, login, forgot password, profile completion (Firebase Auth)
- ☁️ Cloud profiles — education, experience, and skills stored in Firestore
- 🔔 Notifications center
- 🎨 Material Design, Lottie animations, splash screen API
MVVM throughout — each screen is a Fragment/Activity with its own ViewModel and LiveData:
ui/
├── auth/ login, register, forgot-password, complete-profile (+ ViewModels)
├── main/ home, career chat, notifications (+ ViewModels)
├── result/ Gemini call + ranked career results
└── base/ shared activity plumbing
beans/ user, career-path, chat models (Gson)
- AI: Google Generative AI SDK (
gemini-pro) - Backend: Firebase Auth + Firestore
- Navigation: Jetpack Navigation Component
- Logging: Timber
git clone https://github.com/ProHussain/CareerCompass-AI.git- Create a Firebase project (Auth + Firestore) and drop your own
google-services.jsonintoapp/— seegoogle-services.example.jsonfor the expected shape - Get a Gemini API key from Google AI Studio and add to
local.properties:GEMINI_API_KEY=your_key_here - Sync Gradle, run on API 23+
The app intentionally ships no credentials — both Firebase config and the Gemini key are supplied locally.
- Screenshots & demo GIF
- Migrate Gemini calls to the current Google AI SDK + streaming responses
- Kotlin migration
- Repository layer + unit tests for prompt/response parsing
- CI build via GitHub Actions
Ghulam Hussain — Product Engineer · apps with 50M+ combined downloads LinkedIn · GitHub · Medium