diff --git a/README.md b/README.md
index 6749ecb..be9810d 100644
--- a/README.md
+++ b/README.md
@@ -1,10 +1,34 @@
 # iva_mobile
 
-A new Flutter project.
+Voice-to-text mobile client built with Flutter.
 
 ## Getting Started
 
-This project is a starting point for a Flutter application.
+### Dependencies
+
+Audio and speech dependencies live in `pubspec.yaml`:
+
+- `record` – microphone capture and amplitude stream.
+- `speech_to_text` – on-device speech recognition with partial results.
+- `provider` – MVVM state injection.
+
+### Platform permissions
+
+- Android: `android/app/src/main/AndroidManifest.xml` declares
+  `android.permission.RECORD_AUDIO`.
+- iOS: `ios/Runner/Info.plist` includes `NSMicrophoneUsageDescription` and
+  `NSSpeechRecognitionUsageDescription` strings.
+
+### Run locally
+
+```
+flutter pub get
+flutter run
+```
+
+The app boots to the voice screen. Tap the microphone to request permission and
+start recording. The waveform animates from live amplitude data and the
+transcription updates as speech is recognized.
 
 A few resources to get you started if this is your first Flutter project:
 
diff --git a/android/app/src/main/AndroidManifest.xml b/android/app/src/main/AndroidManifest.xml
index 101579e..4b89e27 100644
--- a/android/app/src/main/AndroidManifest.xml
+++ b/android/app/src/main/AndroidManifest.xml
@@ -1,4 +1,5 @@
 <manifest xmlns:android="http://schemas.android.com/apk/res/android">
+    <uses-permission android:name="android.permission.RECORD_AUDIO" />
     <application
         android:label="iva_mobile"
         android:name="${applicationName}"
diff --git a/ios/Runner/Info.plist b/ios/Runner/Info.plist
index f3e11a4..7519746 100644
--- a/ios/Runner/Info.plist
+++ b/ios/Runner/Info.plist
@@ -28,12 +28,16 @@
 	<string>LaunchScreen</string>
 	<key>UIMainStoryboardFile</key>
 	<string>Main</string>
-	<key>UISupportedInterfaceOrientations</key>
-	<array>
-		<string>UIInterfaceOrientationPortrait</string>
-		<string>UIInterfaceOrientationLandscapeLeft</string>
-		<string>UIInterfaceOrientationLandscapeRight</string>
-	</array>
+<key>UISupportedInterfaceOrientations</key>
+<array>
+    <string>UIInterfaceOrientationPortrait</string>
+    <string>UIInterfaceOrientationLandscapeLeft</string>
+    <string>UIInterfaceOrientationLandscapeRight</string>
+</array>
+<key>NSMicrophoneUsageDescription</key>
+<string>This app requires microphone access to capture your voice.</string>
+<key>NSSpeechRecognitionUsageDescription</key>
+<string>Speech recognition is used to transcribe your voice into text.</string>
 	<key>UISupportedInterfaceOrientations~ipad</key>
 	<array>
 		<string>UIInterfaceOrientationPortrait</string>
diff --git a/lib/core/services/audio_capture.dart b/lib/core/services/audio_capture.dart
new file mode 100644
index 0000000..333e00e
--- /dev/null
+++ b/lib/core/services/audio_capture.dart
@@ -0,0 +1,98 @@
+import 'dart:async';
+import 'dart:io';
+
+import 'package:flutter/foundation.dart';
+import 'package:path_provider/path_provider.dart';
+import 'package:record/record.dart';
+
+enum AudioPermissionStatus { granted, denied, restricted }
+
+abstract class AudioCaptureService {
+  Future<AudioPermissionStatus> ensurePermission();
+  Future<void> start();
+  Future<void> pause();
+  Future<void> resume();
+  Future<void> stop();
+  Stream<double> get amplitudeStream; // 0.0 to 1.0
+}
+
+class AudioCaptureServiceImpl implements AudioCaptureService {
+  AudioCaptureServiceImpl({AudioRecorder? recorder})
+    : _recorder = recorder ?? AudioRecorder();
+
+  final AudioRecorder _recorder;
+  final StreamController<double> _levelController =
+      StreamController<double>.broadcast();
+  StreamSubscription<Amplitude>? _subscription;
+
+  @override
+  Stream<double> get amplitudeStream => _levelController.stream;
+
+  @override
+  Future<AudioPermissionStatus> ensurePermission() async {
+    final has = await _recorder.hasPermission();
+    return has ? AudioPermissionStatus.granted : AudioPermissionStatus.denied;
+  }
+
+  @override
+  Future<void> start() async {
+    if (!await _recorder.isRecording()) {
+      final dir = await getTemporaryDirectory();
+      final file = File(
+        '${dir.path}/iva_rec_${DateTime.now().millisecondsSinceEpoch}.m4a',
+      );
+      await _recorder.start(
+        const RecordConfig(encoder: AudioEncoder.aacLc),
+        path: file.path,
+      );
+    }
+    _subscription ??= _recorder
+        .onAmplitudeChanged(const Duration(milliseconds: 80))
+        .listen((amp) {
+          final normalized = _normalizeDb(amp.current);
+          if (!_levelController.isClosed) {
+            _levelController.add(normalized);
+          }
+        });
+  }
+
+  @override
+  Future<void> pause() async {
+    if (await _recorder.isRecording()) {
+      await _recorder.pause();
+    }
+  }
+
+  @override
+  Future<void> resume() async {
+    if (await _recorder.isPaused()) {
+      await _recorder.resume();
+    }
+  }
+
+  @override
+  Future<void> stop() async {
+    try {
+      await _subscription?.cancel();
+    } finally {
+      _subscription = null;
+    }
+    if (await _recorder.isRecording() || await _recorder.isPaused()) {
+      await _recorder.stop();
+    }
+  }
+
+  double _normalizeDb(double db) {
+    // Map [-45dB, 0dB] to [0, 1], clamp outside
+    const minDb = -45.0;
+    const maxDb = 0.0;
+    final clamped = db.clamp(minDb, maxDb);
+    return (clamped - minDb) / (maxDb - minDb);
+  }
+
+  @mustCallSuper
+  void dispose() {
+    _levelController.close();
+    _subscription?.cancel();
+  }
+}
diff --git a/lib/core/services/speech_recognition.dart b/lib/core/services/speech_recognition.dart
new file mode 100644
index 0000000..7d00ce1
--- /dev/null
+++ b/lib/core/services/speech_recognition.dart
@@ -0,0 +1,71 @@
+import 'dart:async';
+
+import 'package:speech_to_text/speech_to_text.dart' as stt;
+
+abstract class SpeechRecognitionService {
+  Future<bool> initialize();
+  Future<void> startListening({bool partialResults = true});
+  Future<void> stopListening();
+  Future<void> cancel();
+  Stream<String> get transcriptionStream; // incremental text
+}
+
+class SpeechRecognitionServiceImpl implements SpeechRecognitionService {
+  SpeechRecognitionServiceImpl({stt.SpeechToText? engine})
+    : _engine = engine ?? stt.SpeechToText();
+
+  final stt.SpeechToText _engine;
+  final StreamController<String> _controller =
+      StreamController<String>.broadcast();
+
+  @override
+  Stream<String> get transcriptionStream => _controller.stream;
+
+  @override
+  Future<bool> initialize() async {
+    final available = await _engine.initialize(
+      onStatus: (_) {},
+      onError: (e) {},
+    );
+    return available;
+  }
+
+  @override
+  Future<void> startListening({bool partialResults = true}) async {
+    if (!_engine.isAvailable) {
+      final ok = await initialize();
+      if (!ok) return;
+    }
+    await _engine.listen(
+      onResult: (result) {
+        final text = result.recognizedWords;
+        if (!_controller.isClosed) {
+          _controller.add(text);
+        }
+      },
+      listenOptions: stt.SpeechListenOptions(
+        listenMode: stt.ListenMode.dictation,
+        partialResults: partialResults,
+        cancelOnError: true,
+      ),
+    );
+  }
+
+  @override
+  Future<void> stopListening() async {
+    if (_engine.isListening) {
+      await _engine.stop();
+    }
+  }
+
+  @override
+  Future<void> cancel() async {
+    if (_engine.isListening) {
+      await _engine.cancel();
+    }
+  }
+
+  void dispose() {
+    _controller.close();
+  }
+}
diff --git a/lib/features/voice_to_text/view/voice_to_text_model.dart b/lib/features/voice_to_text/view/voice_to_text_model.dart
index 71ec762..80e41af 100644
--- a/lib/features/voice_to_text/view/voice_to_text_model.dart
+++ b/lib/features/voice_to_text/view/voice_to_text_model.dart
@@ -1,9 +1,13 @@
 import 'dart:async';
+import 'dart:collection';
 
-import 'package:flutter/material.dart';
+import 'package:flutter/foundation.dart';
+import 'package:iva_mobile/core/services/audio_capture.dart';
+import 'package:iva_mobile/core/services/speech_recognition.dart';
 
 abstract class VoiceToTextModel extends Listenable {
-  List<String> get transcript; // already tokenized words or segments
+  List<String> get transcript; // tokenized words/segments for display
+  String get transcribedText;
   int get activeWordIndex;
   bool get isCursorVisible;
   List<double> get waveformData;
@@ -14,6 +18,7 @@ abstract class VoiceToTextModel extends Listenable {
 
   void setActiveWord(int index);
   void toggleCursorVisibility(bool visible);
+  void updateTranscription(String text, {int? activeWordIndex});
   void updateWaveform(List<double> amplitudes);
   void startTimer();
   void pauseTimer();
@@ -24,24 +29,31 @@ abstract class VoiceToTextModel extends Listenable {
   void stopRecording();
   void restartRecording();
   void discardRecording();
-
-  // Other state already planned (timer, waveform, recording commands) lives here too.
 }
 
 enum RecordingState { idle, recording, paused, stopped }
 
 class VoiceToTextModelState extends ChangeNotifier implements VoiceToTextModel {
   VoiceToTextModelState({
-    required List<String> initialTranscript,
+    List<String> initialTranscript = const [],
     int initialActiveWordIndex = 0,
     bool initialCursorVisible = true,
     RecordingState initialRecordingState = RecordingState.idle,
-  }) : _transcript = List.unmodifiable(initialTranscript),
-       _activeWordIndex = initialActiveWordIndex,
+    this.audio,
+    this.speech,
+    int waveformWindow = 48,
+  }) : _transcriptWords = List<String>.from(initialTranscript),
+       _activeWordIndex = 0,
        _isCursorVisible = initialCursorVisible,
-       _recordingState = initialRecordingState;
+       _recordingState = initialRecordingState,
+       _waveformWindowSize = waveformWindow {
+    _activeWordIndex = _normalizeActiveIndex(initialActiveWordIndex);
+    _seedAmplitudeWindow();
+  }
 
-  final List<String> _transcript;
+  final AudioCaptureService? audio;
+  final SpeechRecognitionService? speech;
+  List<String> _transcriptWords;
   int _activeWordIndex;
   bool _isCursorVisible;
   List<double> _waveformData = const [];
@@ -51,9 +63,18 @@ class VoiceToTextModelState extends ChangeNotifier implements VoiceToTextModel {
   bool _isTimerRunning = false;
   Timer? _timer;
   RecordingState _recordingState;
+  StreamSubscription<double>? _ampSub;
+  StreamSubscription<String>? _sttSub;
+  final List<double> _amplitudeWindow = <double>[];
+  final int _waveformWindowSize;
+  String? _lastError;
+  // No mock fallback: live amplitude must be provided by platform service
+
+  @override
+  List<String> get transcript => UnmodifiableListView(_transcriptWords);
 
   @override
-  List<String> get transcript => _transcript;
+  String get transcribedText => _transcriptWords.join(' ');
 
   @override
   int get activeWordIndex => _activeWordIndex;
@@ -62,7 +83,7 @@ class VoiceToTextModelState extends ChangeNotifier implements VoiceToTextModel {
   bool get isCursorVisible => _isCursorVisible;
 
   @override
-  List<double> get waveformData => _waveformData;
+  List<double> get waveformData => UnmodifiableListView(_waveformData);
 
   @override
   Stream<List<double>> get waveformStream => _waveformController.stream;
@@ -76,11 +97,13 @@ class VoiceToTextModelState extends ChangeNotifier implements VoiceToTextModel {
   @override
   RecordingState get recordingState => _recordingState;
 
+  String? get lastError => _lastError;
+
   @override
   void setActiveWord(int index) {
-    if (index == _activeWordIndex) return;
-    if (index < 0 || index >= _transcript.length) return;
-    _activeWordIndex = index;
+    final normalized = _normalizeActiveIndex(index);
+    if (normalized == _activeWordIndex) return;
+    _activeWordIndex = normalized;
     notifyListeners();
   }
 
@@ -91,63 +114,57 @@ class VoiceToTextModelState extends ChangeNotifier implements VoiceToTextModel {
     notifyListeners();
   }
 
+  @override
+  void updateTranscription(String text, {int? activeWordIndex}) {
+    final words = _normalizeTranscript(text);
+    final changed = _applyTranscription(words, activeWordIndex);
+    if (changed) {
+      notifyListeners();
+    }
+  }
+
   @override
   void updateWaveform(List<double> amplitudes) {
-    _waveformData = List.unmodifiable(amplitudes);
-    notifyListeners();
+    final normalized = amplitudes
+        .map((value) => value.isNaN ? 0.0 : value.clamp(0.0, 1.0))
+        .toList(growable: false);
+    final changed = !listEquals(normalized, _waveformData);
+    _waveformData = List.unmodifiable(normalized);
     if (!_waveformController.isClosed) {
       _waveformController.add(_waveformData);
     }
+    if (changed) {
+      notifyListeners();
+    }
   }
 
   @override
   void startTimer() {
-    if (_isTimerRunning) return;
-    _isTimerRunning = true;
-    _timer ??= Timer.periodic(const Duration(seconds: 1), (_) {
-      _elapsedDuration += const Duration(seconds: 1);
-      notifyListeners();
-    });
-    notifyListeners();
+    _startTimer();
   }
 
   @override
   void pauseTimer() {
-    if (!_isTimerRunning && _timer == null) return;
-    _isTimerRunning = false;
-    _timer?.cancel();
-    _timer = null;
-    notifyListeners();
+    _pauseTimer();
   }
 
   @override
   void resetTimer() {
-    final hadElapsed = _elapsedDuration != Duration.zero;
-    final wasRunning = _isTimerRunning || _timer != null;
-    _elapsedDuration = Duration.zero;
-    if (wasRunning) {
-      _timer?.cancel();
-      _timer = null;
-      _isTimerRunning = false;
-    }
-    if (hadElapsed || wasRunning) {
-      notifyListeners();
-    }
+    _resetTimer();
   }
 
   @override
   void startRecording() {
-    if (_recordingState == RecordingState.recording) return;
-    _recordingState = RecordingState.recording;
-    startTimer();
-    notifyListeners();
+    _beginRecording();
   }
 
   @override
   void pauseRecording() {
     if (_recordingState != RecordingState.recording) return;
     _recordingState = RecordingState.paused;
-    pauseTimer();
+    _pauseTimer(notify: false);
+    audio?.pause();
+    speech?.stopListening();
     notifyListeners();
   }
 
@@ -155,7 +172,9 @@ class VoiceToTextModelState extends ChangeNotifier implements VoiceToTextModel {
   void resumeRecording() {
     if (_recordingState != RecordingState.paused) return;
     _recordingState = RecordingState.recording;
-    startTimer();
+    _startTimer(notify: false);
+    audio?.resume();
+    speech?.startListening(partialResults: true);
     notifyListeners();
   }
 
@@ -163,33 +182,187 @@ class VoiceToTextModelState extends ChangeNotifier implements VoiceToTextModel {
   void stopRecording() {
     if (_recordingState == RecordingState.stopped) return;
     _recordingState = RecordingState.stopped;
-    resetTimer();
+    _resetTimer(notify: false);
+    _teardownStreams();
+    audio?.stop();
+    speech?.stopListening();
     notifyListeners();
   }
 
   @override
   void restartRecording() {
+    _applyTranscription(const [], 0);
+    _seedAmplitudeWindow();
+    _resetTimer(notify: false);
     _recordingState = RecordingState.recording;
-    resetTimer();
-    startTimer();
+    _startTimer(notify: false);
+    _setupStreams();
+    audio?.start();
+    speech?.startListening(partialResults: true);
     notifyListeners();
   }
 
   @override
   void discardRecording() {
+    final transcriptChanged = _applyTranscription(const [], 0);
+    _seedAmplitudeWindow();
+    final timerChanged = _resetTimer(notify: false);
+    _teardownStreams();
+    audio?.stop();
+    speech?.cancel();
+    final previousState = _recordingState;
     _recordingState = RecordingState.idle;
-    _waveformData = const [];
-    if (!_waveformController.isClosed) {
-      _waveformController.add(_waveformData);
+    if (transcriptChanged ||
+        timerChanged ||
+        previousState != RecordingState.idle) {
+      notifyListeners();
     }
-    resetTimer();
-    notifyListeners();
   }
 
   @override
   void dispose() {
     _timer?.cancel();
     _waveformController.close();
+    _teardownStreams();
     super.dispose();
   }
+
+  Future<void> _beginRecording() async {
+    if (_recordingState == RecordingState.recording) {
+      _startTimer(notify: false);
+      return;
+    }
+    if (audio != null) {
+      final status = await audio!.ensurePermission();
+      if (status != AudioPermissionStatus.granted) {
+        _lastError = 'Microphone permission is required to record';
+        notifyListeners();
+        return;
+      }
+    }
+    _recordingState = RecordingState.recording;
+    _startTimer(notify: false);
+    _setupStreams();
+    await audio?.start();
+    await speech?.startListening(partialResults: true);
+    // Web implementation uses getUserMedia; on unsupported env, ensurePermission returns denied
+    notifyListeners();
+  }
+
+  void _setupStreams() {
+    _teardownStreams();
+    if (audio != null) {
+      _ampSub = audio!.amplitudeStream.listen(_pushAmplitude);
+    }
+    if (speech != null) {
+      _sttSub = speech!.transcriptionStream.listen((text) {
+        updateTranscription(text);
+      });
+    }
+  }
+
+  void _teardownStreams() {
+    _ampSub?.cancel();
+    _sttSub?.cancel();
+    _ampSub = null;
+    _sttSub = null;
+    _amplitudeWindow.clear();
+  }
+
+  void _pushAmplitude(double level) {
+    _amplitudeWindow.add(level.clamp(0.0, 1.0));
+    if (_amplitudeWindow.length > _waveformWindowSize) {
+      _amplitudeWindow.removeAt(0);
+    }
+    updateWaveform(List<double>.from(_amplitudeWindow));
+  }
+
+  void _seedAmplitudeWindow() {
+    _amplitudeWindow
+      ..clear()
+      ..addAll(List<double>.filled(_waveformWindowSize, 0.0));
+    updateWaveform(List<double>.from(_amplitudeWindow));
+  }
+
+  // No mock amplitude functions
+
+  bool _applyTranscription(List<String> words, int? activeWordIndex) {
+    final normalizedWords = List<String>.from(words);
+    final normalizedIndex = _normalizeActiveIndex(
+      activeWordIndex,
+      wordCount: normalizedWords.length,
+    );
+    final wordsChanged = !listEquals(normalizedWords, _transcriptWords);
+    final indexChanged = normalizedIndex != _activeWordIndex;
+    if (!wordsChanged && !indexChanged) {
+      return false;
+    }
+    _transcriptWords = normalizedWords;
+    _activeWordIndex = normalizedIndex;
+    return true;
+  }
+
+  bool _startTimer({bool notify = true}) {
+    if (_isTimerRunning) return false;
+    _isTimerRunning = true;
+    _timer ??= Timer.periodic(const Duration(seconds: 1), (_) {
+      _elapsedDuration += const Duration(seconds: 1);
+      notifyListeners();
+    });
+    if (notify) {
+      notifyListeners();
+    }
+    return true;
+  }
+
+  bool _pauseTimer({bool notify = true}) {
+    if (!_isTimerRunning && _timer == null) return false;
+    _isTimerRunning = false;
+    _timer?.cancel();
+    _timer = null;
+    if (notify) {
+      notifyListeners();
+    }
+    return true;
+  }
+
+  bool _resetTimer({bool notify = true}) {
+    final hadElapsed = _elapsedDuration != Duration.zero;
+    final wasRunning = _isTimerRunning || _timer != null;
+    if (!hadElapsed && !wasRunning) {
+      return false;
+    }
+    _elapsedDuration = Duration.zero;
+    if (wasRunning) {
+      _timer?.cancel();
+      _timer = null;
+      _isTimerRunning = false;
+    }
+    if (notify) {
+      notifyListeners();
+    }
+    return true;
+  }
+
+  int _normalizeActiveIndex(int? index, {int? wordCount}) {
+    final total = wordCount ?? _transcriptWords.length;
+    if (total == 0) {
+      return 0;
+    }
+    final desired = index ?? (total - 1);
+    if (desired <= 0) {
+      return 0;
+    }
+    if (desired >= total) {
+      return total - 1;
+    }
+    return desired;
+  }
+
+  List<String> _normalizeTranscript(String text) {
+    return text
+        .split(RegExp(r'\s+'))
+        .where((segment) => segment.isNotEmpty)
+        .toList(growable: false);
+  }
 }
diff --git a/lib/features/voice_to_text/view/voice_to_text_screen.dart b/lib/features/voice_to_text/view/voice_to_text_screen.dart
index 978051f..dbc4ece 100644
--- a/lib/features/voice_to_text/view/voice_to_text_screen.dart
+++ b/lib/features/voice_to_text/view/voice_to_text_screen.dart
@@ -1,8 +1,9 @@
-import 'dart:async';
-import 'dart:math' as math;
+// no-op
 
 import 'package:flutter/material.dart';
 import 'package:provider/provider.dart';
+import 'package:iva_mobile/core/services/audio_capture.dart';
+import 'package:iva_mobile/core/services/speech_recognition.dart';
 
 import '../widget/control_buttons.dart';
 import '../widget/text_display.dart';
@@ -16,7 +17,7 @@ class VoiceToTextScreen extends StatelessWidget {
   @override
   Widget build(BuildContext context) {
     return ChangeNotifierProvider<VoiceToTextModelState>(
-      create: (_) => VoiceToTextModelState(
+      create: (context) => VoiceToTextModelState(
         initialTranscript: const [
           'Herman',
           'is',
@@ -35,6 +36,8 @@ class VoiceToTextScreen extends StatelessWidget {
           'kin',
         ],
         initialActiveWordIndex: 14,
+        audio: context.read<AudioCaptureService>(),
+        speech: context.read<SpeechRecognitionService>(),
       ),
       child: const _VoiceToTextView(),
     );
@@ -49,53 +52,34 @@ class _VoiceToTextView extends StatefulWidget {
 }
 
 class _VoiceToTextViewState extends State<_VoiceToTextView> {
-  static const int _waveformSampleCount = 48;
-  static const Duration _waveformTick = Duration(milliseconds: 100);
-
-  final math.Random _random = math.Random();
-  Timer? _waveformTimer;
+  String? _shownError;
 
   @override
   void initState() {
     super.initState();
-    WidgetsBinding.instance.addPostFrameCallback((_) {
-      final model = context.read<VoiceToTextModelState>();
-      _pushWaveformSample(model);
-      _waveformTimer = Timer.periodic(
-        _waveformTick,
-        (_) => _pushWaveformSample(model),
-      );
-    });
+    // No-op: Real waveform is driven by audio amplitude when recording (issue #7)
   }
 
   @override
   void dispose() {
-    _waveformTimer?.cancel();
     super.dispose();
   }
 
-  void _pushWaveformSample(VoiceToTextModel model) {
-    final halfCount = _waveformSampleCount ~/ 2;
-    final leading = List<double>.generate(halfCount, (index) {
-      final phase = index / halfCount * math.pi;
-      final base = (math.sin(phase) * 0.5) + 0.5;
-      final noise = (_random.nextDouble() - 0.5) * 0.15;
-      return (base + noise).clamp(0.0, 1.0);
-    });
-    final trailing = List<double>.from(leading.reversed);
-    final sample = [...leading, ...trailing];
-    if (sample.length < _waveformSampleCount) {
-      sample.addAll(
-        List<double>.filled(_waveformSampleCount - sample.length, 0.3),
-      );
-    }
-    model.updateWaveform(sample);
-  }
-
   @override
   Widget build(BuildContext context) {
     final VoiceToTextModelState model = context.watch<VoiceToTextModelState>();
 
+    // Report new errors softly via SnackBar
+    WidgetsBinding.instance.addPostFrameCallback((_) {
+      final err = model.lastError;
+      if (err != null && err != _shownError && mounted) {
+        _shownError = err;
+        ScaffoldMessenger.of(
+          context,
+        ).showSnackBar(SnackBar(content: Text(err)));
+      }
+    });
+
     return Scaffold(
       backgroundColor: const Color.fromARGB(255, 227, 227, 198),
       body: SafeArea(
@@ -125,6 +109,7 @@ class _VoiceToTextViewState extends State<_VoiceToTextView> {
                       barWidth: 3,
                       spacing: 6,
                       backgroundColor: Colors.transparent,
+                      minBarHeight: 2,
                     ),
                   ),
                 ),
diff --git a/lib/features/voice_to_text/widget/waveform.dart b/lib/features/voice_to_text/widget/waveform.dart
index c34d0ad..f6a01ce 100644
--- a/lib/features/voice_to_text/widget/waveform.dart
+++ b/lib/features/voice_to_text/widget/waveform.dart
@@ -13,6 +13,7 @@ class WaveformWidget extends StatefulWidget {
     this.spacing = 6,
     this.backgroundColor = Colors.transparent,
     this.animationDuration = const Duration(milliseconds: 120),
+    this.minBarHeight = 2,
   });
 
   final List<double> amplitudes;
@@ -22,6 +23,7 @@ class WaveformWidget extends StatefulWidget {
   final double spacing;
   final Color backgroundColor;
   final Duration animationDuration;
+  final double minBarHeight;
 
   @override
   State<WaveformWidget> createState() => _WaveformWidgetState();
@@ -115,6 +117,7 @@ class _WaveformWidgetState extends State<WaveformWidget>
             barColor: widget.barColor,
             barWidth: widget.barWidth,
             spacing: widget.spacing,
+            minBarHeight: widget.minBarHeight,
           ),
         ),
       ),
@@ -128,12 +131,14 @@ class WaveformPainter extends CustomPainter {
     required this.barColor,
     required this.barWidth,
     required this.spacing,
+    this.minBarHeight = 2,
   });
 
   final List<double> amplitudes;
   final Color barColor;
   final double barWidth;
   final double spacing;
+  final double minBarHeight;
 
   static List<double> normalize(List<double> values) {
     return values
@@ -147,6 +152,20 @@ class WaveformPainter extends CustomPainter {
   @override
   void paint(Canvas canvas, Size size) {
     if (amplitudes.isEmpty) {
+      // Draw a flat baseline using minBarHeight
+      final paint = Paint()
+        ..color = barColor
+        ..strokeWidth = barWidth
+        ..strokeCap = StrokeCap.round;
+      final centerY = size.height / 2;
+      final half = minBarHeight / 2;
+      final totalWidth = barWidth + spacing; // minimal placeholder segment
+      final startX = (size.width - totalWidth) / 2;
+      canvas.drawLine(
+        Offset(startX, centerY - half),
+        Offset(startX, centerY + half),
+        paint,
+      );
       return;
     }
 
@@ -162,7 +181,8 @@ class WaveformPainter extends CustomPainter {
 
     for (var i = 0; i < amplitudes.length; i++) {
       final normalized = amplitudes[i].clamp(0.0, 1.0).toDouble();
-      final barHeight = normalized * maxHeight;
+      final barHeight =
+          (normalized * (maxHeight - minBarHeight)) + minBarHeight;
       final x = startX + i * (barWidth + spacing);
       final top = (maxHeight - barHeight) / 2;
       canvas.drawLine(Offset(x, top), Offset(x, top + barHeight), paint);
@@ -189,6 +209,7 @@ class WaveformStream extends StatelessWidget {
     this.spacing = 6,
     this.backgroundColor = Colors.transparent,
     this.animationDuration = const Duration(milliseconds: 120),
+    this.minBarHeight = 2,
   });
 
   final Stream<List<double>> stream;
@@ -199,6 +220,7 @@ class WaveformStream extends StatelessWidget {
   final double spacing;
   final Color backgroundColor;
   final Duration animationDuration;
+  final double minBarHeight;
 
   @override
   Widget build(BuildContext context) {
@@ -215,6 +237,7 @@ class WaveformStream extends StatelessWidget {
           spacing: spacing,
           backgroundColor: backgroundColor,
           animationDuration: animationDuration,
+          minBarHeight: minBarHeight,
         );
       },
     );
diff --git a/lib/main.dart b/lib/main.dart
index eed1679..e7d7068 100644
--- a/lib/main.dart
+++ b/lib/main.dart
@@ -1,5 +1,8 @@
 import 'package:flutter/material.dart';
+import 'package:provider/provider.dart';
 
+import 'core/services/audio_capture.dart';
+import 'core/services/speech_recognition.dart';
 import 'features/voice_to_text/view/voice_to_text_screen.dart';
 
 void main() {
@@ -11,15 +14,23 @@ class VoiceToTextApp extends StatelessWidget {
 
   @override
   Widget build(BuildContext context) {
-    return MaterialApp(
-      debugShowCheckedModeBanner: false,
-      title: 'IVA Voice',
-      theme: ThemeData(
-        colorScheme: ColorScheme.fromSeed(seedColor: Colors.black87),
-        scaffoldBackgroundColor: const Color(0xFFE3E3C6),
-        useMaterial3: true,
+    return MultiProvider(
+      providers: [
+        Provider<AudioCaptureService>(create: (_) => AudioCaptureServiceImpl()),
+        Provider<SpeechRecognitionService>(
+          create: (_) => SpeechRecognitionServiceImpl(),
+        ),
+      ],
+      child: MaterialApp(
+        debugShowCheckedModeBanner: false,
+        title: 'IVA Voice',
+        theme: ThemeData(
+          colorScheme: ColorScheme.fromSeed(seedColor: Colors.black87),
+          scaffoldBackgroundColor: const Color(0xFFE3E3C6),
+          useMaterial3: true,
+        ),
+        home: const VoiceToTextScreen(),
       ),
-      home: const VoiceToTextScreen(),
     );
   }
 }
diff --git a/linux/flutter/generated_plugin_registrant.cc b/linux/flutter/generated_plugin_registrant.cc
index e71a16d..9209285 100644
--- a/linux/flutter/generated_plugin_registrant.cc
+++ b/linux/flutter/generated_plugin_registrant.cc
@@ -6,6 +6,10 @@
 
 #include "generated_plugin_registrant.h"
 
+#include <record_linux/record_linux_plugin.h>
 
 void fl_register_plugins(FlPluginRegistry* registry) {
+  g_autoptr(FlPluginRegistrar) record_linux_registrar =
+      fl_plugin_registry_get_registrar_for_plugin(registry, "RecordLinuxPlugin");
+  record_linux_plugin_register_with_registrar(record_linux_registrar);
 }
diff --git a/linux/flutter/generated_plugins.cmake b/linux/flutter/generated_plugins.cmake
index 2e1de87..29e96ee 100644
--- a/linux/flutter/generated_plugins.cmake
+++ b/linux/flutter/generated_plugins.cmake
@@ -3,6 +3,7 @@
 #
 
 list(APPEND FLUTTER_PLUGIN_LIST
+  record_linux
 )
 
 list(APPEND FLUTTER_FFI_PLUGIN_LIST
diff --git a/macos/Flutter/GeneratedPluginRegistrant.swift b/macos/Flutter/GeneratedPluginRegistrant.swift
index cccf817..d63568b 100644
--- a/macos/Flutter/GeneratedPluginRegistrant.swift
+++ b/macos/Flutter/GeneratedPluginRegistrant.swift
@@ -5,6 +5,12 @@
 import FlutterMacOS
 import Foundation
 
+import path_provider_foundation
+import record_darwin
+import speech_to_text_macos
 
 func RegisterGeneratedPlugins(registry: FlutterPluginRegistry) {
+  PathProviderPlugin.register(with: registry.registrar(forPlugin: "PathProviderPlugin"))
+  RecordPlugin.register(with: registry.registrar(forPlugin: "RecordPlugin"))
+  SpeechToTextMacosPlugin.register(with: registry.registrar(forPlugin: "SpeechToTextMacosPlugin"))
 }
diff --git a/pubspec.lock b/pubspec.lock
index 8f3e021..a48033f 100644
--- a/pubspec.lock
+++ b/pubspec.lock
@@ -121,6 +121,14 @@ packages:
       url: "https://pub.dev"
     source: hosted
     version: "1.3.3"
+  ffi:
+    dependency: transitive
+    description:
+      name: ffi
+      sha256: "289279317b4b16eb2bb7e271abccd4bf84ec9bdcbe999e278a94b804f5630418"
+      url: "https://pub.dev"
+    source: hosted
+    version: "2.1.4"
   file:
     dependency: transitive
     description:
@@ -129,6 +137,14 @@ packages:
       url: "https://pub.dev"
     source: hosted
     version: "7.0.1"
+  fixnum:
+    dependency: transitive
+    description:
+      name: fixnum
+      sha256: b6dc7065e46c974bc7c5f143080a6764ec7a4be6da1285ececdc37be96de53be
+      url: "https://pub.dev"
+    source: hosted
+    version: "1.1.1"
   flutter:
     dependency: "direct main"
     description: flutter
@@ -152,6 +168,11 @@ packages:
     description: flutter
     source: sdk
     version: "0.0.0"
+  flutter_web_plugins:
+    dependency: transitive
+    description: flutter
+    source: sdk
+    version: "0.0.0"
   freezed_annotation:
     dependency: transitive
     description:
@@ -288,6 +309,78 @@ packages:
       url: "https://pub.dev"
     source: hosted
     version: "1.9.1"
+  path_provider:
+    dependency: "direct main"
+    description:
+      name: path_provider
+      sha256: "50c5dd5b6e1aaf6fb3a78b33f6aa3afca52bf903a8a5298f53101fdaee55bbcd"
+      url: "https://pub.dev"
+    source: hosted
+    version: "2.1.5"
+  path_provider_android:
+    dependency: transitive
+    description:
+      name: path_provider_android
+      sha256: "993381400e94d18469750e5b9dcb8206f15bc09f9da86b9e44a9b0092a0066db"
+      url: "https://pub.dev"
+    source: hosted
+    version: "2.2.18"
+  path_provider_foundation:
+    dependency: transitive
+    description:
+      name: path_provider_foundation
+      sha256: "16eef174aacb07e09c351502740fa6254c165757638eba1e9116b0a781201bbd"
+      url: "https://pub.dev"
+    source: hosted
+    version: "2.4.2"
+  path_provider_linux:
+    dependency: transitive
+    description:
+      name: path_provider_linux
+      sha256: f7a1fe3a634fe7734c8d3f2766ad746ae2a2884abe22e241a8b301bf5cac3279
+      url: "https://pub.dev"
+    source: hosted
+    version: "2.2.1"
+  path_provider_platform_interface:
+    dependency: transitive
+    description:
+      name: path_provider_platform_interface
+      sha256: "88f5779f72ba699763fa3a3b06aa4bf6de76c8e5de842cf6f29e2e06476c2334"
+      url: "https://pub.dev"
+    source: hosted
+    version: "2.1.2"
+  path_provider_windows:
+    dependency: transitive
+    description:
+      name: path_provider_windows
+      sha256: bd6f00dbd873bfb70d0761682da2b3a2c2fccc2b9e84c495821639601d81afe7
+      url: "https://pub.dev"
+    source: hosted
+    version: "2.3.0"
+  pedantic:
+    dependency: transitive
+    description:
+      name: pedantic
+      sha256: "67fc27ed9639506c856c840ccce7594d0bdcd91bc8d53d6e52359449a1d50602"
+      url: "https://pub.dev"
+    source: hosted
+    version: "1.11.1"
+  platform:
+    dependency: transitive
+    description:
+      name: platform
+      sha256: "5d6b1b0036a5f331ebc77c850ebc8506cbc1e9416c27e59b439f917a902a4984"
+      url: "https://pub.dev"
+    source: hosted
+    version: "3.1.6"
+  plugin_platform_interface:
+    dependency: transitive
+    description:
+      name: plugin_platform_interface
+      sha256: "4820fbfdb9478b1ebae27888254d445073732dae3d6ea81f0b7e06d5dedc3f02"
+      url: "https://pub.dev"
+    source: hosted
+    version: "2.1.8"
   provider:
     dependency: "direct main"
     description:
@@ -312,6 +405,70 @@ packages:
       url: "https://pub.dev"
     source: hosted
     version: "1.5.0"
+  record:
+    dependency: "direct main"
+    description:
+      name: record
+      sha256: "9dbc6ff3e784612f90a9b001373c45ff76b7a08abd2bd9fdf72c242320c8911c"
+      url: "https://pub.dev"
+    source: hosted
+    version: "6.1.1"
+  record_android:
+    dependency: transitive
+    description:
+      name: record_android
+      sha256: "854627cd78d8d66190377f98477eee06ca96ab7c9f2e662700daf33dbf7e6673"
+      url: "https://pub.dev"
+    source: hosted
+    version: "1.4.2"
+  record_ios:
+    dependency: transitive
+    description:
+      name: record_ios
+      sha256: "13e241ed9cbc220534a40ae6b66222e21288db364d96dd66fb762ebd3cb77c71"
+      url: "https://pub.dev"
+    source: hosted
+    version: "1.1.2"
+  record_linux:
+    dependency: transitive
+    description:
+      name: record_linux
+      sha256: "235b1f1fb84e810f8149cc0c2c731d7d697f8d1c333b32cb820c449bf7bb72d8"
+      url: "https://pub.dev"
+    source: hosted
+    version: "1.2.1"
+  record_macos:
+    dependency: transitive
+    description:
+      name: record_macos
+      sha256: "2849068bb59072f300ad63ed146e543d66afaef8263edba4de4834fc7c8d4d35"
+      url: "https://pub.dev"
+    source: hosted
+    version: "1.1.1"
+  record_platform_interface:
+    dependency: transitive
+    description:
+      name: record_platform_interface
+      sha256: b0065fdf1ec28f5a634d676724d388a77e43ce7646fb049949f58c69f3fcb4ed
+      url: "https://pub.dev"
+    source: hosted
+    version: "1.4.0"
+  record_web:
+    dependency: transitive
+    description:
+      name: record_web
+      sha256: "4f0adf20c9ccafcc02d71111fd91fba1ca7b17a7453902593e5a9b25b74a5c56"
+      url: "https://pub.dev"
+    source: hosted
+    version: "1.2.0"
+  record_windows:
+    dependency: transitive
+    description:
+      name: record_windows
+      sha256: "223258060a1d25c62bae18282c16783f28581ec19401d17e56b5205b9f039d78"
+      url: "https://pub.dev"
+    source: hosted
+    version: "1.0.7"
   sky_engine:
     dependency: transitive
     description: flutter
@@ -325,6 +482,38 @@ packages:
       url: "https://pub.dev"
     source: hosted
     version: "1.10.1"
+  speech_to_text:
+    dependency: "direct main"
+    description:
+      name: speech_to_text
+      sha256: c07557664974afa061f221d0d4186935bea4220728ea9446702825e8b988db04
+      url: "https://pub.dev"
+    source: hosted
+    version: "7.3.0"
+  speech_to_text_platform_interface:
+    dependency: transitive
+    description:
+      name: speech_to_text_platform_interface
+      sha256: a1935847704e41ee468aad83181ddd2423d0833abe55d769c59afca07adb5114
+      url: "https://pub.dev"
+    source: hosted
+    version: "2.3.0"
+  speech_to_text_windows:
+    dependency: transitive
+    description:
+      name: speech_to_text_windows
+      sha256: "2c9846d18253c7bbe059a276297ef9f27e8a2745dead32192525beb208195072"
+      url: "https://pub.dev"
+    source: hosted
+    version: "1.0.0+beta.8"
+  sprintf:
+    dependency: transitive
+    description:
+      name: sprintf
+      sha256: "1fc9ffe69d4df602376b52949af107d8f5703b77cda567c4d7d86a0693120f23"
+      url: "https://pub.dev"
+    source: hosted
+    version: "7.0.0"
   stack_trace:
     dependency: transitive
     description:
@@ -373,6 +562,14 @@ packages:
       url: "https://pub.dev"
     source: hosted
     version: "1.4.0"
+  uuid:
+    dependency: transitive
+    description:
+      name: uuid
+      sha256: a5be9ef6618a7ac1e964353ef476418026db906c4facdedaa299b7a2e71690ff
+      url: "https://pub.dev"
+    source: hosted
+    version: "4.5.1"
   vector_math:
     dependency: transitive
     description:
@@ -397,6 +594,22 @@ packages:
       url: "https://pub.dev"
     source: hosted
     version: "1.1.3"
+  web:
+    dependency: transitive
+    description:
+      name: web
+      sha256: "868d88a33d8a87b18ffc05f9f030ba328ffefba92d6c127917a2ba740f9cfe4a"
+      url: "https://pub.dev"
+    source: hosted
+    version: "1.1.1"
+  xdg_directories:
+    dependency: transitive
+    description:
+      name: xdg_directories
+      sha256: "7a3f37b05d989967cdddcbb571f1ea834867ae2faa29725fd085180e0883aa15"
+      url: "https://pub.dev"
+    source: hosted
+    version: "1.1.0"
   yaml:
     dependency: transitive
     description:
@@ -407,4 +620,4 @@ packages:
     version: "3.1.3"
 sdks:
   dart: ">=3.9.2 <4.0.0"
-  flutter: ">=3.18.0-18.0.pre.54"
+  flutter: ">=3.29.0"
diff --git a/pubspec.yaml b/pubspec.yaml
index 05df03c..409205c 100644
--- a/pubspec.yaml
+++ b/pubspec.yaml
@@ -33,6 +33,9 @@ dependencies:
   provider: ^6.1.5+1
   flutter_localizations:
     sdk: flutter
+  record: ^6.1.1
+  speech_to_text: ^7.3.0
+  path_provider: ^2.1.5
 
   # The following adds the Cupertino Icons font to your application.
   # Use with the CupertinoIcons class for iOS style icons.
diff --git a/test/features/voice_to_text/view/voice_to_text_model_test.dart b/test/features/voice_to_text/view/voice_to_text_model_test.dart
index 978dca5..e46c1ff 100644
--- a/test/features/voice_to_text/view/voice_to_text_model_test.dart
+++ b/test/features/voice_to_text/view/voice_to_text_model_test.dart
@@ -14,11 +14,13 @@ void main() {
         initialCursorVisible: false,
       );
 
-      expect(model.transcript, transcript);
+      expect(model.transcript, equals(transcript));
       expect(() => model.transcript.add('extra'), throwsUnsupportedError);
       expect(model.activeWordIndex, 1);
       expect(model.isCursorVisible, isFalse);
-      expect(model.waveformData, isEmpty);
+      expect(model.waveformData, isNotEmpty);
+      expect(model.waveformData.every((v) => v == 0.0), isTrue);
+      expect(model.transcribedText, 'Herman is recording');
     });
 
     test('updates active word and notifies listeners', () {
@@ -36,7 +38,7 @@ void main() {
       expect(notifyCount, 1);
     });
 
-    test('ignores out of range indices', () {
+    test('clamps out of range indices', () {
       final model = VoiceToTextModelState(initialTranscript: transcript);
       var notifyCount = 0;
       model.addListener(() => notifyCount++);
@@ -44,8 +46,8 @@ void main() {
       model.setActiveWord(-1);
       model.setActiveWord(transcript.length);
 
-      expect(model.activeWordIndex, 0);
-      expect(notifyCount, 0);
+      expect(model.activeWordIndex, transcript.length - 1);
+      expect(notifyCount, 1);
     });
 
     test('toggles cursor visibility and notifies once per change', () {
@@ -77,11 +79,11 @@ void main() {
 
         model.updateWaveform([0.1, 0.5, 1.2]);
 
-        expect(model.waveformData, [0.1, 0.5, 1.2]);
+        expect(model.waveformData, equals([0.1, 0.5, 1.0]));
         expect(() => model.waveformData.add(0.3), throwsUnsupportedError);
         await Future<void>.delayed(Duration.zero);
         expect(events, [
-          [0.1, 0.5, 1.2],
+          [0.1, 0.5, 1.0],
         ]);
         expect(notifyCount, 1);
 
@@ -133,6 +135,25 @@ void main() {
       });
     });
 
+    test('updateTranscription replaces transcript and active word', () {
+      final model = VoiceToTextModelState(initialTranscript: transcript);
+      var notifyCount = 0;
+      model.addListener(() => notifyCount++);
+
+      model.updateTranscription('Hello world from iva', activeWordIndex: 1);
+
+      expect(model.transcript, equals(['Hello', 'world', 'from', 'iva']));
+      expect(model.transcribedText, 'Hello world from iva');
+      expect(model.activeWordIndex, 1);
+      expect(notifyCount, 1);
+
+      model.updateTranscription('Hello world from iva', activeWordIndex: 1);
+      expect(notifyCount, 1, reason: 'no change, no notification');
+
+      model.updateTranscription('Hello world from iva');
+      expect(model.activeWordIndex, 3);
+    });
+
     test('recording state transitions control timer lifecycle', () {
       fakeAsync((async) {
         final model = VoiceToTextModelState(initialTranscript: transcript);
@@ -161,15 +182,19 @@ void main() {
         model.restartRecording();
         expect(model.recordingState, RecordingState.recording);
         expect(model.elapsedDuration, Duration.zero);
+        expect(model.transcript, isEmpty);
 
         async.elapse(const Duration(seconds: 1));
         expect(model.elapsedDuration, const Duration(seconds: 1));
 
+        model.updateTranscription('partial text');
         model.updateWaveform([0.2, 0.4]);
         model.discardRecording();
         expect(model.recordingState, RecordingState.idle);
         expect(model.elapsedDuration, Duration.zero);
-        expect(model.waveformData, isEmpty);
+        expect(model.waveformData, isNotEmpty);
+        expect(model.waveformData.every((v) => v == 0.0), isTrue);
+        expect(model.transcript, isEmpty);
 
         model.dispose();
       });
diff --git a/windows/flutter/generated_plugin_registrant.cc b/windows/flutter/generated_plugin_registrant.cc
index 8b6d468..458cde3 100644
--- a/windows/flutter/generated_plugin_registrant.cc
+++ b/windows/flutter/generated_plugin_registrant.cc
@@ -6,6 +6,9 @@
 
 #include "generated_plugin_registrant.h"
 
+#include <record_windows/record_windows_plugin_c_api.h>
 
 void RegisterPlugins(flutter::PluginRegistry* registry) {
+  RecordWindowsPluginCApiRegisterWithRegistrar(
+      registry->GetRegistrarForPlugin("RecordWindowsPluginCApi"));
 }
diff --git a/windows/flutter/generated_plugins.cmake b/windows/flutter/generated_plugins.cmake
index b93c4c3..e9670a9 100644
--- a/windows/flutter/generated_plugins.cmake
+++ b/windows/flutter/generated_plugins.cmake
@@ -3,6 +3,7 @@
 #
 
 list(APPEND FLUTTER_PLUGIN_LIST
+  record_windows
 )
 
 list(APPEND FLUTTER_FFI_PLUGIN_LIST