diff --git a/app.py b/app.py
index d82c51f0d..3b987398c 100644
--- a/app.py
+++ b/app.py
@@ -1,6 +1,55 @@
-from flask import Flask
+from flask import Flask, request
+import os
+import openai
+
 app = Flask(__name__)
+openai.api_key = 'YOUR_OPEN_AI_KEY'
 
 @app.route('/')
-def hello_world():
-    return 'Hello, World!'
+def index():
+    return """
+    <html>
+        <body>
+            <button onclick="startRecording()">Start Recording</button>
+            <button onclick="stopRecording()">Stop Recording</button>
+            <script>
+                let rec;
+                let audioChunks = [];
+                
+                function startRecording() {
+                    audioChunks = [];
+                    rec = new MediaRecorder(window.stream);
+                    rec.ondataavailable = e => {
+                        audioChunks.push(e.data);
+                    };
+                    rec.start();
+                }
+
+                function stopRecording() {
+                    rec.stop();
+                    const audioBlob = new Blob(audioChunks);
+                    const reader = new FileReader();
+                    reader.readAsDataURL(audioBlob);
+                    reader.onloadend = () => {
+                        const base64data = reader.result;
+                        fetch('/transcribe', {
+                            method: 'POST',
+                            body: JSON.stringify({ data: base64data }),
+                            headers: { 'Content-Type': 'application/json' }
+                        }).then(response => response.text()).then(data => {
+                            console.log(data);
+                        });
+                    };
+                }
+            </script>
+        </body>
+    </html>
+    """
+
+@app.route('/transcribe', methods=['POST'])
+def transcribe_audio():
+    audio_data = request.json['data']
+    # Here you need to convert the base64 audio data into a format suitable for the OpenAI API.
+    # This will likely involve writing the data to a file, or converting it to the correct audio format.
+    transcript = openai.Audio.transcribe("whisper-1", audio_data)
+    return transcript
