**With Papo (https://papo.global/), you can use artificial intelligence as a co-pilot for your customer service team, suggesting the best responses in real-time during a conversation. Through modules like Capitão, the platform analyzes the customer's question and recommends relevant answers, either from a list of predefined Quick Replies or by generating text based on your Knowledge Base documents, increasing the speed and consistency of human support.**

Efficiency in customer service depends on quick and accurate responses. However, even the most experienced agents can take time to find the correct information or type out answers to repetitive questions. Papo's agent-assist AI solves this problem by acting as an assistant that reads the conversation and offers response suggestions with a single click.

This allows human agents to focus on customer relationships, while technology handles the repetitive task of finding and formulating information.

## Types of AI-Generated Suggestions

Papo's AI uses two main sources to suggest responses to your agents during customer service. Each has a specific purpose and, together, they create a robust support system.

| Suggestion Type | Information Source | Ideal For |
| :--- | :--- | :--- |
| **Quick Replies** | Pre-approved texts manually created by your team. | Very frequent and standardized questions, such as greetings, prices, operating hours, useful links, and exchange policies. |
| **Knowledge Base** | Documents (PDFs, product manuals, internal policies) that you upload to the platform. | More complex or specific questions that require consulting detailed information not found in a ready-made response. |

Quick Replies ensure standardization and speed for common queries, while the Knowledge Base offers depth and flexibility for unexpected questions.

## How AI response suggestion works

The process is designed to be intuitive and not interrupt the agent's workflow. Papo's Capitão module functions as a co-pilot integrated into the conversation screen.

1.  **Customer Message Analysis:** When a customer sends a question, the AI analyzes the content in real-time to understand the intent behind the message.
2.  **Search for the Best Response:** Based on this analysis, the AI searches on two fronts:
    *   Checks if the question matches any of the configured **Quick Replies**.
    *   Consults the **Knowledge Base** to find document excerpts that answer the customer's query.
3.  **Suggestion Presentation:** The platform displays the most relevant suggestions directly in the agent interface, next to the typing box. Suggestions can include both Quick Replies and texts generated from documents.
4.  **Agent Action:** The agent can then:
    *   Click on a suggestion to insert it directly into the conversation.
    *   Edit the suggestion before sending it to personalize the message.
    *   Ignore the suggestions and type a manual response, if none are suitable.

This cycle happens in seconds, transforming customer service into a more dynamic and efficient process.

## How to configure AI to suggest responses

For AI to assist your team, you first need to provide it with the necessary knowledge. Configuration is divided into simple steps and does not require advanced technical knowledge.

| Step | Action | Description |
| :--- | :--- | :--- |
| **1. Feed the Knowledge Base** | Upload relevant documents (PDFs, text files) to the "Knowledge Base" section of Papo. | Include product manuals, company policies, troubleshooting guides, and frequently asked questions (FAQs) lists. |
| **2. Create Quick Replies** | Access the "Quick Replies" area and create shortcuts and texts for the most common questions. | Organize them into folders by department (Sales, Support) to facilitate management and use by agents. |
| **3. Activate the AI Assistant** | In your AI assistant (Capitão) settings, link the Knowledge Bases and Quick Replies it should use. | You can have different assistants for different teams, each with its own set of knowledge. |
| **4. Train the Team** | Show agents how suggestions will appear on the conversation screen and how they can use them. | Usage is intuitive, but a brief alignment ensures everyone gets the most out of the functionality. |

## Frequently Asked Questions (FAQ)

### What is the difference between AI that responds autonomously and AI that suggests responses?
At Papo, there are two main applications of AI. **Super Chat** is an autonomous AI agent that responds to customers 24/7, qualifies leads, and schedules appointments without human intervention. **Capitão**, on the other hand, is an AI co-pilot that *assists* the human agent, suggesting responses during the conversation so the agent can work faster.

### Where does AI get the information to suggest responses?
AI retrieves information from two locations that you configure: the **Quick Replies** library, which contains pre-approved texts for common questions, and the **Knowledge Base**, composed of documents (such as PDFs and manuals) that you upload to the platform.

### Can the agent edit an AI-suggested response before sending it?
Yes. AI suggestions are a starting point. The agent has full control and can click on the suggestion to insert it into the text box, and then edit, add, or remove information to personalize the message before sending it to the customer.

### Do I need technical knowledge to configure the Knowledge Base?
No. The process involves uploading files your company already has, such as product manuals, PDF return policies, or text documents with FAQs. The platform automatically processes these files to make them searchable by AI.

### Are Quick Replies the same as what chatbots use?
Not exactly. Quick Replies are primarily a tool for human agents, used to streamline manual customer service. The chatbot's AI (Super Chat) primarily uses the Knowledge Base to formulate its own responses autonomously, although both functionalities can be fed by similar information sources.

### Does AI learn from the responses agents choose?
AI is trained to identify the best response based on the content you provide in the Knowledge Base and Quick Replies. The effectiveness of suggestions is directly linked to the quality and organization of this information. The platform does not use agent choices to autonomously retrain the model, ensuring that control over the information remains with you.

## Read also
*   [How to use Quick Replies to speed up customer service](/hc/papo/articles/how-to-use-quick-replies-to-streamline-customer-service)
*   [What are AI Agents and how do they work at Papo?](/hc/papo/articles/what-are-ai-agents-and-how-they-work-in-papo)

---

**[Start using for free →](https://papo.global/?utm_source=helpcenter&utm_medium=article&utm_campaign=how-to-use-ai-to-suggest-best-response-during-customer-service&utm_content=en)**

The first user is free, and you can start configuring your customer service operation right away.