TwinMind2

Watch Demo

Watch the demo video on Google Drive:
▶️ View TwinMind2 Demo

An intelligent Android meeting recorder application powered by Google Gemini AI that provides real-time transcription, intelligent summaries, and comprehensive meeting insights.

Features

🎙️ Smart Recording

• Continuous Audio Recording: High-quality audio capture with automatic chunking
• Pause/Resume: Control recording with pause and resume functionality
• Silence Detection: Automatically detects if microphone is working
• Phone Call Integration: Automatically pauses recording during incoming calls

📝 Real-time Transcription

• Auto-Transcription: Automatically transcribes audio chunks using Google Gemini 2.0 Flash
• Progress Tracking: Real-time status updates for transcription progress
• Retry Logic: Automatic retries with exponential backoff on failures
• Separate Transcript View: Dedicated screen for viewing complete transcripts

📊 AI-Powered Summaries

• Intelligent Summarization: Generate comprehensive meeting summaries using Gemini AI
• Meeting Title: Auto-generate concise meeting titles
• Action Items: Extract and list action items from discussions
• Key Points: Identify and highlight important decisions and topics
• Separate Summary View: Dedicated screen for viewing summaries
• Progressive Generation: Watch summaries generate section by section

🎯 Session Management

• Audio Sessions: Organize recordings by session
• Complete Audio: Access full audio recordings of entire sessions
• Chunk Viewing: View individual audio chunks
• Status Tracking: Track recording, transcription, and summary status

⚠️ IMPORTANT SETUP NOTE

Before running the app, you MUST configure your Google Gemini API key!

The app requires a valid Google Gemini API key to function. Follow the setup instructions below to add your key.

Setup Instructions

Prerequisites

• Android Studio Hedgehog (2023.1.1) or later
• JDK 11 or later
• Android SDK 24 (Android 7.0) or higher
• Google Gemini API key (Get one here)

Installation

1. Clone the repository

   git clone <repository-url>
   cd TwinMind2

2. ⚠️ Configure API Key - REQUIRED STEP

   DO THIS BEFORE BUILDING: You must add your Google Gemini API key to run the app!

   Edit the file app/src/main/java/com/example/twinmind2/di/NetworkModule.kt and replace YOUR_GOOGLE_GEMINI_API_KEY_HERE with your actual API key:

   fun getGeminiApiKey(context: android.content.Context): String {
       // Replace with your actual Google Gemini API key
       val hardcodedApiKey = "YOUR_GOOGLE_GEMINI_API_KEY_HERE"  // ⬅️ CHANGE THIS!
       
       // Try to get from SharedPreferences first, fallback to hardcoded
       val prefs = context.getSharedPreferences("twinmind_prefs", android.content.Context.MODE_PRIVATE)
       val savedKey = prefs.getString("gemini_api_key", "")?.takeIf { it.isNotEmpty() }
       
       return savedKey ?: hardcodedApiKey
   }

   Get your API key from: Google AI Studio

   • Click "Create API Key"
   • Copy the generated key
   • Paste it in NetworkModule.kt where indicated above

3. Build the project

   ./gradlew assembleDebug

4. Run the app

   • Open the project in Android Studio
   • Connect an Android device or start an emulator
   • Click Run (Shift+F10)

Architecture

Tech Stack

• Language: Kotlin
• UI Framework: Jetpack Compose
• Architecture: MVVM with Hilt Dependency Injection
• Database: Room Database
• Network: Retrofit, OkHttp
• AI: Google Gemini 2.0 Flash
• Navigation: Jetpack Navigation Compose

Key Components

• MainActivity: Main UI with recording controls and session navigation
• RecordingService: Foreground service handling audio recording and transcription
• RecordingViewModel: ViewModel managing recording state
• TranscriptionRepository: Handles Gemini API transcription calls
• SummaryRepository: Manages AI-powered summary generation
• RecordingRepository: Manages audio chunks and session data

Database Schema

• RecordingSession: Stores session metadata (start/end times, status, complete audio path)
• AudioChunk: Individual 30-second audio segments with overlap for continuity
• Transcript: Transcribed text for each chunk
• Summary: AI-generated meeting summaries with title, summary, action items, and key points

Usage

Recording a Meeting

1. Launch the app
2. Tap Record to start recording
3. Use Pause during breaks or Resume to continue
4. Tap Stop when the meeting ends
5. View transcripts and summaries automatically

Viewing Transcripts

1. Tap on any session to expand it
2. Click View Transcript button
3. See real-time transcription progress
4. Read the complete transcript

Generating Summaries

1. Navigate to a session
2. Click View Summary button
3. If not generated, tap Generate Summary
4. Watch as sections generate:
   • Meeting Title
   • Summary
   • Action Items
   • Key Points

Error Handling

API Rate Limiting (429)

The app automatically handles Google Gemini API rate limits:

• Automatic Retries: Up to 5 retries with exponential backoff
• Wait Times: 2s, 4s, 8s, 16s, 32s between retries
• User Feedback: Clear error messages if retries fail

Network Issues

• Retry Logic: Built-in retry mechanisms for failed API calls
• Progress Tracking: Real-time status updates during failures
• Offline Capability: Audio chunks saved locally for processing when online

API Key Issues

If you encounter API key errors:

1. Verify your API key is valid at Google AI Studio
2. Check that API is enabled in Google Cloud Console
3. Ensure billing is enabled for your Google Cloud project
4. Update the key in NetworkModule.kt

Permissions

The app requires the following permissions:

• RECORD_AUDIO: Capture audio from microphone
• READ_PHONE_STATE: Detect incoming calls to pause recording
• FOREGROUND_SERVICE: Run background recording service
• FOREGROUND_SERVICE_MICROPHONE: Foreground microphone access
• POST_NOTIFICATIONS: Show recording status notifications
• INTERNET: Connect to Gemini API
• ACCESS_NETWORK_STATE: Check network connectivity

File Storage

Audio files are stored in the app's internal storage:

• Location: /data/data/com.example.twinmind2/files/recordings/{sessionId}/
• Format: WAV files (16-bit PCM, 16kHz, mono)
• Chunk files: chunk_XXXX.wav (30-second segments with 2-second overlap)
• Complete audio: complete_audio.wav (full session concatenated)

Limitations

• Recording Duration: Limited by device storage
• API Limits: Subject to Google Gemini API quotas
• Storage: Audio files can consume significant storage space
• Network: Requires internet for transcription and summarization

Contributing

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Submit a pull request

License

This project is licensed under the MIT License.

Acknowledgments

• Google Gemini AI: For transcription and summarization capabilities
• Android Architecture Components: For modern app architecture
• Jetpack Compose: For declarative UI framework

Version

Current Version: 1.0