app.py

import gradio as gr
from STT.sst import speech_to_text
from LLM.llm import generate_reply
from TTS_X.tts import generate_voice

def process(audio):

    user_text = speech_to_text(audio)
    reply = generate_reply(user_text)
    reply_audio = generate_voice(reply)
    
    return user_text, reply, reply_audio

with gr.Blocks() as demo:
    gr.Markdown("## 🗣️➡️💬➡️🔊 من صوتك إلى رد منطوق!")

    audio_input = gr.Audio(label="🎤 ارفع صوتك", type="filepath")
    user_text = gr.Textbox(label="📜 النص المسموع")
    reply_text = gr.Textbox(label="🤖 رد المساعد")
    reply_audio = gr.Audio(label="🔊 الرد بالصوت")

    btn = gr.Button("ابدأ")

    btn.click(process, inputs=audio_input, outputs=[user_text, reply_text, reply_audio])

demo.launch()