File size: 7,245 Bytes
ea6dab8
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179


import os
import json
import base64
import asyncio
from fastapi import FastAPI, Request, Response
from twilio.twiml.voice_response import VoiceResponse, Gather, Hangup
from twilio.rest import Client
import openai
from elevenlabs.client import ElevenLabs
import logging
from logging.handlers import RotatingFileHandler

# --- Configuration ---
# Secrets are loaded from environment variables, not a config file.
TWILIO_ACCOUNT_SID = os.environ.get('TWILIO_ACCOUNT_SID')
TWILIO_AUTH_TOKEN = os.environ.get('TWILIO_AUTH_TOKEN')
OPENAI_API_KEY = os.environ.get('OPENAI_API_KEY')
ELEVENLABS_API_KEY = os.environ.get('ELEVENLABS_API_KEY')
ELEVENLABS_VOICE_ID = 'Xb7hH8MSUjPSbSDYk0k2'  # Hardcoded Voice ID for "Alice"

# Setup clients
twilio_client = Client(TWILIO_ACCOUNT_SID, TWILIO_AUTH_TOKEN)
openai.api_key = OPENAI_API_KEY
elevenlabs_client = ElevenLabs(api_key=ELEVENLABS_API_KEY)

# Setup logging
log_formatter = logging.Formatter('%(asctime)s - %(levelname)s - [%(filename)s:%(lineno)d] - %(message)s')
log_handler = RotatingFileHandler('app.log', maxBytes=10*1024*1024, backupCount=5)
log_handler.setFormatter(log_formatter)

logger = logging.getLogger(__name__)
logger.setLevel(logging.INFO)
logger.addHandler(log_handler)

app = FastAPI()

# In-memory conversation state
conversations = {}

@app.post("/")
async def handle_incoming_call(request: Request):
    """Handle incoming calls and start the conversation."""
    form = await request.form()
    call_sid = form.get('CallSid', 'Unknown')
    user_language = form.get('lang', 'en-US') # Default to English
    logger.info(f"Incoming call {call_sid} in {user_language}")

    # Initialize conversation history
    conversations[call_sid] = {"history": [], "language": user_language}

    response = VoiceResponse()
    try:
        with open('sales_script.txt', 'r') as f:
            sales_script = f.read()

        # Generate initial greeting
        completion = openai.chat.completions.create(
            model="gpt-4o",
            messages=[
                {"role": "system", "content": f"You are a world-class AI sales agent for bracketworks.io. Your goal is to sell credit cards. Start the conversation with a friendly greeting in {user_language}."},
                {"role": "user", "content": sales_script}
            ]
        )
        greeting = completion.choices[0].message.content
        conversations[call_sid]["history"].append({"role": "assistant", "content": greeting})

        # Generate audio for the greeting
        audio_stream = elevenlabs_client.generate(
            text=greeting,
            voice=ELEVENLABS_VOICE_ID,
            model="eleven_multilingual_v2",
            stream=True
        )
        audio_base64 = base64.b64encode(b"".join(audio_stream)).decode('utf-8')
        response.play(f"data:audio/mp3;base64,{audio_base64}")

    except Exception as e:
        logger.error(f"Error in initial greeting: {e}")
        response.say("Hello, thank you for calling. I am having some technical difficulties.")

    gather = Gather(input='speech', action='/gather', speechTimeout='auto', bargeIn=True, language=user_language)
    response.append(gather)
    return Response(content=str(response), media_type="application/xml")

@app.post("/gather")
async def gather_speech(request: Request):
    """Handle user speech and generate AI response."""
    form = await request.form()
    call_sid = form.get('CallSid', 'Unknown')
    speech_text = form.get('SpeechResult', '').lower()
    user_language = form.get('language', conversations.get(call_sid, {}).get('language', 'en-US'))
    logger.info(f"Speech from {call_sid}: '{speech_text}' in {user_language}")

    response = VoiceResponse()

    if "bye" in speech_text:
        response.say("Thank you for your time. Goodbye.")
        response.hangup()
        return Response(content=str(response), media_type="application/xml")

    if call_sid in conversations:
        conversations[call_sid]["history"].append({"role": "user", "content": speech_text})
        conversations[call_sid]["language"] = user_language # Update language if it changes

    try:
        with open('sales_script.txt', 'r') as f:
            sales_script = f.read()

        completion = openai.chat.completions.create(
            model="gpt-4o",
            messages=[
                {"role": "system", "content": f"You are a persuasive AI sales agent for bracketworks.io. Your goal is to sell credit cards based on the following script: {sales_script}. Respond in {user_language}."},
            ] + conversations.get(call_sid, {}).get("history", []),
            stream=True
        )
        ai_response_text = ""
        for chunk in completion:
            ai_response_text += chunk.choices[0].delta.content or ""

        conversations.get(call_sid, {}).get("history", []).append({"role": "assistant", "content": ai_response_text})

        audio_stream = elevenlabs_client.generate(
            text=ai_response_text,
            voice=ELEVENLABS_VOICE_ID,
            model="eleven_multilingual_v2",
            stream=True
        )
        audio_base64 = base64.b64encode(b"".join(audio_stream)).decode('utf-8')
        response.play(f"data:audio/mp3;base64,{audio_base64}")

    except Exception as e:
        logger.error(f"Error during AI response generation: {e}")
        response.say("I am sorry, I am having trouble processing your request.")

    gather = Gather(input='speech', action='/gather', speechTimeout='auto', bargeIn=True, language=user_language)
    response.append(gather)
    return Response(content=str(response), media_type="application/xml")

@app.post("/status")
async def call_status(request: Request):
    """Handle call status updates and data extraction."""
    form = await request.form()
    call_sid = form.get('CallSid', 'Unknown')
    call_status = form.get('CallStatus', 'Unknown')
    logger.info(f"Call {call_sid} status: {call_status}")

    if call_status == 'completed' and call_sid in conversations:
        try:
            history = conversations[call_sid]["history"]
            conversation_text = "\n".join([f"{msg['role']}: {msg['content']}" for msg in history])

            completion = openai.chat.completions.create(
                model="gpt-4o",
                messages=[
                    {"role": "system", "content": 'Extract Name, Phone Number, and Email from the conversation. Also provide a brief summary. Respond in JSON format with keys: "name", "phone", "email", "summary".'},
                    {"role": "user", "content": conversation_text}
                ]
            )
            extracted_data = json.loads(completion.choices[0].message.content)

            if not os.path.exists('application_data'):
                os.makedirs('application_data')
            with open(f"application_data/{call_sid}.json", "w") as f:
                json.dump(extracted_data, f)

            logger.info(f"Saved extracted data for {call_sid}")
        except Exception as e:
            logger.error(f"Error extracting data for {call_sid}: {e}")
        finally:
            del conversations[call_sid]

    return Response(status_code=200)

if __name__ == '__main__':
    import uvicorn
    uvicorn.run(app, host="0.0.0.0", port=8080)