> For clean Markdown of any page, append `.md` to the page URL.
> For a complete documentation index, see https://docs.sarvam.ai/llms.txt.
> For full documentation content in one file, see https://docs.sarvam.ai/llms-full.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.sarvam.ai/_mcp/server.

# Collection Agent using LiveKit

> Build a voice-based collection agent for payment reminders and follow-ups using LiveKit and Sarvam AI. Support for 11 languages (10 Indian + English) with natural voices.

## Overview

This guide demonstrates how to build a **voice-based collection agent** that can handle payment reminders, follow-ups, and payment assistance using **LiveKit** for real-time communication and **Sarvam AI** for speech processing. Perfect for fintech companies, banks, and lending institutions serving Indian customers.

> **Tip**
>
> For the broader architecture pattern this agent fits into, telephony ingress, escalation, latency targets, see the [IVR & Contact Center](/api/api-guides-tutorials/speech-to-text/use-cases/ivr-contact-center) use-case guide.

## What You'll Build

A collection agent that can:

* Make professional payment reminder calls in multiple Indian languages
* Handle customer queries about payments, due dates, and payment options
* Guide customers through payment processes
* Maintain a professional and empathetic tone

## Quick Overview

1. Get API keys (LiveKit, Sarvam)
2. Install packages
3. Create `.env` file with your API keys
4. Write the agent code
5. Run: `python agent.py dev`
6. Test: `python agent.py console`

---

## Quick Start

### 1. Prerequisites

* Python 3.9 or higher
* API keys from:
  * [LiveKit Cloud](https://cloud.livekit.io) (free account)
  * [Sarvam AI](https://dashboard.sarvam.ai) (get API key from dashboard)

### 2. Install Dependencies

#### macOS/Linux

```bash
pip install "livekit-agents[sarvam,silero]" python-dotenv
```

#### Windows

```bash
pip install livekit-agents[sarvam,silero] python-dotenv
```

### 3. Create Environment File

Create a file named `.env` in your project folder and add your API keys:

```env
LIVEKIT_URL=wss://your-project-xxxxx.livekit.cloud
LIVEKIT_API_KEY=APIxxxxxxxxxxxxx
LIVEKIT_API_SECRET=xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
SARVAM_API_KEY=sk_xxxxxxxxxxxxxxxxxxxxxxxx
```

Replace the values with your actual API keys.

### 4. Write Your Agent

Create `collection_agent.py`:

```python
import logging
from dotenv import load_dotenv
from livekit.agents import JobContext, WorkerOptions, cli
from livekit.agents.voice import Agent, AgentSession
from livekit.plugins import sarvam

# Load environment variables
load_dotenv()

# Set up logging
logger = logging.getLogger("collection-agent")
logger.setLevel(logging.INFO)


class CollectionAgent(Agent):
    def __init__(self) -> None:
        super().__init__(
            # Collection agent personality and instructions
            instructions="""
                You are a professional and empathetic collection agent working for ABC Bank.
                
                Customer Account Details:
                - Bank Name: ABC Bank
                - EMI Amount: ₹5,000
                - Due Date: 15th January 2025
                - Loan Type: Personal Loan
                - Account Status: Payment Overdue
                
                Your responsibilities:
                - Remind customers about their pending EMI payment of ₹5,000 which was due on 15th January
                - Provide information about payment due dates, amounts, and available payment methods
                - Help customers understand their payment options and any applicable late fees
                - Guide customers through the payment process if they want to pay immediately
                - Address customer concerns about their account with empathy
                - Offer payment plans or extensions when appropriate (mention that you can connect them with a specialist)
                
                Payment Methods to mention:
                - UPI payment to ABC Bank
                - Net Banking
                - ABC Bank mobile app
                - Visit nearest ABC Bank branch
                
                Communication guidelines:
                - Always maintain a professional yet friendly tone
                - Be empathetic to customer's financial situations
                - Never be aggressive, threatening, or use inappropriate language
                - If a customer is upset, remain calm and understanding
                - Speak clearly and concisely
                - Confirm important details like EMI amount (₹5,000) and due date (15th January)
                - If customer requests to speak to a human, acknowledge and offer to transfer
                
                Start by greeting the customer, introducing yourself as calling from ABC Bank, 
                and politely remind them about their pending EMI of ₹5,000.
            """,
            
            # Saaras v4 STT - Converts speech to text
            stt=sarvam.STT(
                language="unknown",  # Auto-detect language
                model="saaras:v4",
                mode="transcribe"
            ),
            
            # Sarvam LLM - The "brain" that processes and generates responses
            llm=sarvam.LLM(model="sarvam-105b"),
            
            # Bulbul TTS - Converts text to speech
            tts=sarvam.TTS(
                language_code="en-IN",
                model="bulbul:v3",
                speaker="aditya"  # Professional male voice
            ),
        )
    
    async def on_enter(self):
        """Called when user joins - agent starts the conversation"""
        self.session.generate_reply()


async def entrypoint(ctx: JobContext):
    """Main entry point - LiveKit calls this when a user connects"""
    logger.info(f"User connected to room: {ctx.room.name}")
    
    # Create and start the agent session
    session = AgentSession()
    await session.start(
        agent=CollectionAgent(),
        room=ctx.room
    )


if __name__ == "__main__":
    # Run the agent
    cli.run_app(WorkerOptions(entrypoint_fnc=entrypoint))
```

### 5. Run Your Agent

```bash
python collection_agent.py dev
```

### 6. Test Your Agent

In a new terminal, run:

```bash
python collection_agent.py console
```

---

## Customization Examples

### Example 1: Hindi Collection Agent

For customers who prefer Hindi:

```python
stt=sarvam.STT(
    language="hi-IN",  # Hindi
    model="saaras:v4",
    mode="transcribe"
),
tts=sarvam.TTS(
    language_code="hi-IN",
    model="bulbul:v3",
    speaker="anand"  # Professional Hindi male voice
)
```

### Example 2: Tamil Collection Agent

```python
stt=sarvam.STT(language="ta-IN", model="saaras:v4", mode="transcribe"),
tts=sarvam.TTS(
    language_code="ta-IN",
    model="bulbul:v3",
    speaker="priya"
)
```

### Example 3: Multilingual Agent (Auto-detect)

```python
stt=sarvam.STT(language="unknown", model="saaras:v4", mode="transcribe"),  # Auto-detects language
tts=sarvam.TTS(language_code="en-IN", model="bulbul:v3", speaker="aditya")
```

---

## Available Options

### Language Codes

| Language        | Code      |
| --------------- | --------- |
| English (India) | `en-IN`   |
| Hindi           | `hi-IN`   |
| Bengali         | `bn-IN`   |
| Tamil           | `ta-IN`   |
| Telugu          | `te-IN`   |
| Gujarati        | `gu-IN`   |
| Kannada         | `kn-IN`   |
| Malayalam       | `ml-IN`   |
| Marathi         | `mr-IN`   |
| Punjabi         | `pa-IN`   |
| Odia            | `od-IN`   |
| Auto-detect     | `unknown` |

### Speaker Voices (Bulbul v3)

**Male (23):** Shubh (default), Aditya, Rahul, Rohan, Amit, Dev, Ratan, Varun, Manan, Sumit, Kabir, Aayan, Ashutosh, Advait, Anand, Tarun, Sunny, Mani, Gokul, Vijay, Mohit, Rehan, Soham

**Female (14):** Ritu, Priya, Neha, Pooja, Simran, Kavya, Ishita, Shreya, Roopa, Tanya, Shruti, Suhani, Kavitha, Rupali

---

## Pro Tips

* Use `language="unknown"` to automatically detect the language - great for diverse customer bases
* Use a professional male voice like `aditya` or `anand` for collection calls
* Sarvam's models handle code-mixing naturally - customers can switch between languages mid-conversation
* Always maintain compliance with collection regulations in your jurisdiction

---

## Troubleshooting

**API key errors**: Check that all keys are in your `.env` file and the file is in the same directory as your script.

**Module not found**: Run the installation command again based on your operating system.

**Poor transcription**: Try `language="unknown"` for auto-detection, or specify the correct language code.

---

## Additional Resources

* [Sarvam AI Documentation](https://docs.sarvam.ai)
* [LiveKit Documentation](https://docs.livekit.io)
* [LiveKit Sarvam LLM Plugin](https://docs.livekit.io/agents/models/llm/sarvam/)
* [LiveKit Sarvam STT Plugin](https://docs.livekit.io/agents/models/stt/plugins/sarvam/)
* [LiveKit Sarvam TTS Plugin](https://docs.livekit.io/agents/models/tts/plugins/sarvam/)

---

## Need Help?

* Sarvam Support: [support@sarvam.ai](mailto:support@sarvam.ai)
* Community: [Join the Discord Community](https://discord.com/invite/5rAsykttcs)

---

**Happy Building!**

> Build a voice-based collection agent for payment reminders and follow-ups using LiveKit and Sarvam AI. Support for 11 languages (10 Indian + English) with natural voices.