diff --git a/.gitignore b/.gitignore
new file mode 100644
index 0000000..4c49bd7
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1 @@
+.env
diff --git a/requirements.txt b/requirements.txt
index 742213c..fbc9097 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -1,3 +1,8 @@
 fastapi==0.115.2
 uvicorn==0.31.1
 httpx==0.27.2
+elevenlabs==2.18.0
+python-dotenv==1.1.1
+fastf1==3.6.1
+pandas==2.3.3
+requests==2.32.5
\ No newline at end of file
diff --git a/scripts/data/audio/pit_call.mp3 b/scripts/data/audio/pit_call.mp3
new file mode 100644
index 0000000..8e4f633
Binary files /dev/null and b/scripts/data/audio/pit_call.mp3 differ
diff --git a/scripts/fetch_race_data.py b/scripts/fetch_race_data.py
old mode 100644
new mode 100755
index ed8769b..1d5a1a9
--- a/scripts/fetch_race_data.py
+++ b/scripts/fetch_race_data.py
@@ -175,7 +175,7 @@ def prepare_telemetry_stream(telemetry: pd.DataFrame, sample_rate_hz: float = 10
     # Resample to target rate if needed
     telemetry = telemetry.copy()
     telemetry['Time'] = pd.to_timedelta(telemetry['Time'])
-    telemetry = telemetry.sort_values('Time')
+    telemetry = telemetry.sort_values(['LapNumber', 'Time'])
     
     # Convert to milliseconds for easier time tracking
     telemetry['TimeMs'] = (telemetry['Time'].dt.total_seconds() * 1000).astype(int)
diff --git a/scripts/strategy_voice_integration.py b/scripts/strategy_voice_integration.py
new file mode 100644
index 0000000..0c39dae
--- /dev/null
+++ b/scripts/strategy_voice_integration.py
@@ -0,0 +1,31 @@
+"""
+Example: Integrate voice feedback with AI strategy decisions
+"""
+
+from voice_service import RaceEngineerVoice
+from pathlib import Path
+
+def announce_strategy_decision(decision: dict):
+    """
+    Convert AI strategy decision to voice announcement.
+    
+    Args:
+        decision: Dict with keys like 'action', 'tire_compound', 'lap'
+    """
+    engineer = RaceEngineerVoice()
+    
+    # Generate appropriate message
+    if decision['action'] == 'pit':
+        text = f"Box this lap for {decision['tire_compound']}. In, in, in!"
+    elif decision['action'] == 'stay_out':
+        text = "Stay out. These tires are still competitive"
+    elif decision['action'] == 'push':
+        text = f"Push mode. We need {decision.get('gap_target', 3)} seconds"
+    else:
+        text = decision.get('message', 'Copy that')
+    
+    # Synthesize and save
+    audio_path = Path(f"data/audio/lap_{decision.get('lap', 0)}_command.mp3")
+    engineer.synthesize_strategy_message(text, audio_path)
+    
+    return audio_path
\ No newline at end of file
diff --git a/scripts/voice_service.py b/scripts/voice_service.py
new file mode 100644
index 0000000..d30454a
--- /dev/null
+++ b/scripts/voice_service.py
@@ -0,0 +1,90 @@
+"""
+ElevenLabs Voice Service for F1 Race Engineer
+
+Converts AI strategy recommendations to natural speech.
+"""
+
+import os
+from pathlib import Path
+from elevenlabs.client import ElevenLabs
+from elevenlabs import save
+from dotenv import load_dotenv
+
+load_dotenv()
+
+class RaceEngineerVoice:
+    def __init__(self, voice_id: str = "mbBupyLcEivjpxh8Brkf"):  # Default: Rachel
+        """
+        Initialize ElevenLabs voice service.
+        
+        Args:
+            voice_id: ElevenLabs voice ID (Rachel is default, professional female voice)
+        """
+        self.client = ElevenLabs(api_key=os.getenv("ELEVENLABS_API_KEY"))
+        self.voice_id = voice_id
+        
+    def synthesize_strategy_message(
+        self, 
+        text: str, 
+        output_path: Path,
+        stability: float = 0.4,
+        similarity_boost: float = 0.95
+    ) -> Path:
+        """
+        Convert strategy text to speech.
+        
+        Args:
+            text: Message to synthesize (e.g., "Box this lap, box this lap")
+            output_path: Where to save audio file
+            stability: Voice stability (0-1)
+            similarity_boost: Voice similarity (0-1)
+        
+        Returns:
+            Path to generated audio file
+        """
+        audio = self.client.text_to_speech.convert(
+            voice_id=self.voice_id,
+            text=text,
+            model_id="eleven_multilingual_v2",  # Fast, low-latency model
+            voice_settings={
+                "stability": stability,
+                "similarity_boost": similarity_boost,
+                "style": 0.7,
+                "use_speaker_boost": True
+            }
+        )
+        
+        # Save audio
+        output_path = Path(output_path)
+        output_path.parent.mkdir(parents=True, exist_ok=True)
+        
+        # Write the audio generator to file
+        save(audio, str(output_path))
+        
+        return output_path
+    
+    def race_engineer_commands(self) -> dict:
+        """Common F1 race engineer commands"""
+        return {
+            "box_now": "Box this lap, box this lap",
+            "stay_out": "Stay out, stay out. We're looking good on these tires",
+            "push": "Push now, push. We need to build a gap",
+            "save_tires": "Manage those tires. Lift and coast into the corners",
+            "traffic_ahead": "Traffic ahead. Blue flags expected",
+            "safety_car": "Safety car, safety car. We're checking strategy",
+            "undercut_threat": "Undercut threat from behind. We may need to respond",
+            "fastest_lap": "You're on for fastest lap. Push this lap",
+        }
+
+
+# Example usage
+if __name__ == "__main__":
+    engineer = RaceEngineerVoice()
+    
+    # Generate pit call
+    audio_file = engineer.synthesize_strategy_message(
+        text="Mama Mia! That was a close one! Let's keep going and finish this out strong, like a cheetah on espresso!",
+        output_path=Path("data/audio/pit_call.mp3")
+    )
+    
+    print(f"✓ Generated: {audio_file}")
\ No newline at end of file