💬 Responses on Behalf of the Bot

The neuro-employee responds itself. But sometimes you want to intervene: provide a more accurate answer, resolve an unusual situation, or simply take the initiative into your own hands.

For this, there is the response on behalf of the bot feature.

---

How It Works

When a client messages the neuro-employee (in Telegram, in the widget on the website, or in any connected channel), you can:

1. Open the control panel.
2. See the conversation.
3. Write a response — it will be sent on behalf of the bot.

The client sees no difference. For them, the communication is with the same employee. There’s no need to write “an operator is responding to you” or switch to another account.

---

In Telegram

The same feature works in the Telegram bot.

When a user messages @aiworkersbot or your company’s connected bot, you can:

• see the entire conversation in real-time;
• respond on behalf of the bot — the message will reach the client as a response from the assistant;
• take the initiative if the dialogue enters a complex topic;
• after your response, the neuro-employee continues working in other dialogues.

---

On the Website (Widget)

The same logic applies on the website.

A visitor opens the widget and starts a dialogue. The assistant responds. If you see that the answer requires personal involvement:

• open the control panel;
• write a response on behalf of the bot;
• the message goes to the visitor as a response from the assistant;
• the visitor continues the conversation, unaware that a person just responded.

---

Why This is Necessary

• Unconventional questions that are not in the knowledge base.
• Complex claims that require human evaluation.
• Individual conditions that cannot be left to an automated response.
• Live control — you see that the neuro-employee is responding, and you can join at any moment.

Meanwhile, the neuro-employee continues to respond independently in other dialogues — you do not block its work.

---

How This Differs from Passing it to a Human

With handing off to a human, the neuro-employee pauses the dialogue — adds ! to the username and waits for an operator to take control.

With responses on behalf of the bot, the intervention is seamless: the neuro-employee seems to "respond," but a person writes for them. The dialogue does not break.

---

Key Points

Responding on behalf of the bot is a way to control the quality of communication without switching context for the client.

You can:

• respond from Telegram or the website;
• intervene in any dialogue at any moment;
• remain "behind the scenes" — the client communicates with the same employee.

Learn more: Widget for the Website.