Update app.py

app.py

@@ -18,51 +18,63 @@ Answer questions about your background, experience, projects, and skills based o
 """
 
 # Chat function that takes a question + API key and keeps history
-def chat(user_input, history, api_key):
+def transcribe_and_chat(audio, history, api_key):
     if not api_key:
-        return "❌ Please enter your OpenAI API key.", history
+        return history, "❌ Please enter your OpenAI API key."
 
     try:
         client = OpenAI(api_key=api_key)
-        messages = [{"role": "system", "content": system_prompt}]
-
-        # Append chat history
-        for user_msg, bot_msg in history:
-            messages.append({"role": "user", "content": user_msg})
-            messages.append({"role": "assistant", "content": bot_msg})
 
-        messages.append({"role": "user", "content": user_input})
-
-        response = client.chat.completions.create(
+        # 1) Transcribe the audio file with Whisper
+        with open(audio, "rb") as f:
+            whisper_resp = client.audio.transcriptions.create(
+                model="whisper-1",
+                file=f
+            )
+        user_text = whisper_resp.text.strip()
+        
+        # 2) Append user message to history
+        messages = [{"role": "system", "content": system_prompt}]
+        for u, b in history:
+            messages += [
+                {"role": "user",      "content": u},
+                {"role": "assistant", "content": b}
+            ]
+        messages.append({"role": "user", "content": user_text})
+
+        # 3) ChatCompletion
+        chat_resp = client.chat.completions.create(
             model="gpt-4",
-            temperature=0.7,
-            messages=messages
+            messages=messages,
+            temperature=0.7
         )
+        bot_reply = chat_resp.choices[0].message.content.strip()
 
-        bot_reply = response.choices[0].message.content.strip()
-        history.append((user_input, bot_reply))
-        return bot_reply, history
+        # 4) Update history
+        history.append((user_text, bot_reply))
+        return history, history
 
     except Exception as e:
-        return f"❌ Error: {str(e)}", history
+        return history, f"❌ Error: {e}"
 
-# Gradio interface
 with gr.Blocks(title="Voice Bot: Krishnavamshi Thumma") as demo:
-    gr.Markdown("## 🎙️ Ask me about my experience, projects, or skills")
+    gr.Markdown("## 🎙️ Speak your question; get a text answer")
 
-    api_key = gr.Textbox(label="🔐 Enter your OpenAI API Key", type="password")
+    api_key = gr.Textbox(label="🔐 OpenAI API Key", type="password")
     chatbot = gr.Chatbot(label="🧠 Voice Bot", type="messages")
-    msg = gr.Textbox(label="💬 Ask a question")
-    clear = gr.Button("Clear")
-
-    history = gr.State([])
-
-    def respond(message, history, key):
-        reply, updated = chat(message, history, key)
-        return updated, updated
-
-    msg.submit(respond, [msg, history, api_key], [chatbot, history])
-    clear.click(lambda: ([], []), None, [chatbot, history])
-
-# Launch the Gradio app
-demo.launch()
+    mic      = gr.Audio(source="microphone", type="filepath", label="🎤 Record your question")
+    clear    = gr.Button("Clear chat")
+    state    = gr.State([])
+
+    # When mic component gets new audio, run transcription + chat
+    mic.change(
+        fn=transcribe_and_chat, 
+        inputs=[mic, state, api_key], 
+        outputs=[chatbot, state]
+    )
+
+    # Clear history
+    clear.click(lambda: ([], []), None, [chatbot, state])
+
+    # Note: we don’t need share=True on Spaces
+    demo.launch()