ChannelPulse AI is a Flutter-powered, AI-driven application that dynamically adapts onboarding flows and business dashboards based on a user’s source channel. Leveraging Gemini-style intent prediction, it personalizes everything — from UI layout and CTAs to AI-generated storytelling insights.
“Feels like the app was made just for them.” – Your users, after using ChannelPulse AI
🎥 Watch the Demo Video
📲 Try the App
Inspired by Gemini AI and FlutterFlow UX best practices, ChannelPulse AI reimagines user onboarding and insights delivery for growth-oriented teams and digital products.
| Category | Description |
|---|---|
| 🔍 Source Detection | Detects user source via UTM, Firebase Dynamic Links, or deep linking |
| 🎯 Intent Prediction | Uses Gemini-style prediction to determine user persona (e.g., buyer, explorer) |
| 🧩 Adaptive UI | Tailors onboarding and CTA copy based on predicted intent |
| 📈 AI Narrative Dashboard | Real-time, AI-generated insights with storytelling visuals and charts |
| 🔊 Voice Script (Optional) | Gemini-generated voice narration for accessible insights (TTS) |
| 📤 Executive Report Export | One-click PDF summary of key insights for stakeholders |
- Frontend: Flutter (Dart)
- Backend: Gemini API (mocked or real), Firebase (optional)
- State Management: Flutter
Navigator,FutureBuilder, conditional rendering - Charts:
syncfusion_flutter_chartsorfl_chart(recommended)
lib/ ├── main.dart # App entry point ├── pages/ │ ├── landing_page.dart # Detect & select traffic source │ ├── dynamic_onboarding_page.dart # Adaptive onboarding flow │ └── ai_insights_dashboard.dart # Narrative insight dashboard ├── services/ │ └── gemini_service.dart # Intent prediction integration ├── models/ │ └── insight.dart # Data model for AI insights assets/ └── banners/ # Custom illustrations per user intent
-
Clone the Repository
git clone https://github.com/your-username/channelpulse-flutter.git cd channelpulse-flutter -
Install Dependencies flutter pub get
-
Run the App flutter run
Optional: Connect Firebase Dynamic Links or pass UTM parameters for realistic simulation.
🤖 Gemini API (Mock or Real) You can simulate Gemini-style responses by editing gemini_service.dart, or integrate with a real Gemini-compatible backend (Flask/FastAPI).
Sample Response Format
{ "intent": "highly motivated buyer", "heroTitle": "Start your fitness journey today!", "ctaButton": "Join the Movement", "tone": "Casual, energetic" }
✨ Example Use Cases
| Channel Source | Predicted Intent | UI Behavior |
|---|---|---|
| High visual engagement | Minimal text, colorful UI, vibrant CTA | |
| Professional research intent | Informative layout, trust-building tone | |
| Referral | Ready to signup | Direct signup, strong CTA like “Get Started Now” |
📊 AI-Powered Dashboard Smart Cards: Summarized KPIs with contextual highlights
Interactive Charts: Visualize data trends with fl_chart / syncfusion_flutter_charts
Narrated Insights: Voice-over summaries (TTS enabled)
Export Summary: Generate PDF/CSV insight reports with one click
📄 License This project is licensed under the MIT License.
👩💻 Author Sudarshanam Yessasvini 📧 [email protected] 🌐 Portfolio 🔗 LinkedIn
💡 Contributing PRs and suggestions are welcome! If you find a bug or want to add new Gemini-powered features, feel free to open an issue or fork the repo.
⭐️ Show Your Support If you like this project, give it a ⭐️ and share it with your AI/Flutter community!