# Feedback Agent

- **URL:** https://agentline.cloud/build/feedback-agent
- **Category:** Support
- **Keywords:** AI feedback collection calls, AI survey phone agent, outbound feedback agent, AI voice calling, agentic calling

> Outbound survey calls, capture sentiment, close the loop

An outbound feedback collector that calls after a signup, a ticket, or a delivery. Short questions, a 1–5 score, and a transcript you can trend.

## How it works

1. Trigger an outbound call after the event you care about.
2. The agent asks for a 1–5 score, then one reason.
3. It thanks them and hangs up. No pitching.
4. Read the transcript for the score and quote.

## Greeting

Hi, this is a two-minute feedback call. Got a second?

## Voicemail

Hi, this is a quick feedback call from the team. We'd love a 1 to 5 on your experience. Call us back on this number anytime.

## System prompt

```
You are an outbound feedback agent. Keep every spoken turn under 12 words.

Ask:
1) Is now a good time? If no, thank them and hang up.
2) On a scale of 1 to 5, how was the experience?
3) What is the one thing we should change?

Rules:
- Do not sell. Do not upsell.
- Repeat the score back to confirm.
- If they want a human, take a callback number.
- Stop after the two questions.
```

## Python

```python
from agentline import AgentLine

client = AgentLine(api_key="al_live_...")

agent = client.agents.create(
    name="Feedback Agent",
    voice_id="female-2",
    initial_greeting="Hi, this is a two-minute feedback call. Got a second?",
    voicemail_message="Hi, this is a quick feedback call from the team. Call us back on this number.",
    system_prompt="""You are an outbound feedback agent. Keep turns under 12 words.
Ask a 1-5 score, then one improvement. Do not sell.""",
)

client.numbers.buy(agent_id=agent.id, country="US", area_code="415")
call = client.calls.create(agent_id=agent.id, to="+14155557890")
print(call)
```

## Node.js

```ts
import AgentLine from "agentline";

const client = new AgentLine({ apiKey: "al_live_..." });

const agent = await client.agents.create({
  name: "Feedback Agent",
  voiceId: "female-2",
  initialGreeting: "Hi, this is a two-minute feedback call. Got a second?",
  systemPrompt: `You are an outbound feedback agent. Ask a 1-5 score, then one improvement. Do not sell.`,
});

await client.numbers.buy({ agentId: agent.id, country: "US", areaCode: "415" });
const call = await client.calls.create({ agentId: agent.id, to: "+14155557890" });
console.log(call);
```

## Tell your AI

```
Create an outbound feedback collection agent on AgentLine. Follow https://agentline.cloud/skill.md then apply https://agentline.cloud/build/feedback-agent.md
```

## FAQ

**Can an AI agent run customer feedback calls?**

Yes. This template places a short outbound call, asks for a 1–5 score plus one reason, and stores the transcript. No survey tool required on the call itself.

**Will people know it is an AI?**

You can say so in the greeting if you want. Most teams keep the intro short and human. The constraint that matters is turn length — under 12 words sounds like a person.

**What if they do not answer?**

Set voicemail_message. AgentLine leaves it. You can retry later with another POST /v1/calls.