AI Agent Dashboard Overview

The AI Agent Dashboard is where you create, manage, and monitor your AI agents in Sense. 

From this single screen, you can view all your agents, search and filter them by status, use case, or channel, and manage their configurations, knowledge base articles, and fallback responses to keep candidate conversations accurate and up to date.

Navigating to Agent Builder

You can follow the steps below to navigate to the Agent Builder:

1. Log in to the Sense platform.
2. Go to the main navigation menu from the left menu and click on Automation.
3. Click on AI Agents.
4. You will land on the AI Agent page where you can create or manage agents.

AI Agents Screen Overview

The All Agents screen shows all the agents created in your account in one place. You can use this screen to view, search, and manage your agents, as well as create new ones.

Each agent listed here includes details like its Name, Entity Type, Use Case, Last Updated, Created Owner, Channels & Status, helping you quickly understand and manage your agents.

Upon initial access, the screen is empty, presenting suggestions to initiate the creation of a new agent. Once agents are created, they will be displayed in this area.

Search & Filter

You can use either the search bar or filters to narrow down the list of agents based on specific criteria.

• Status → View agents based on their current state (for example: Draft, Ready to use, In Use or Archived)
• Use Case → Filter agents based on their purpose (for example: Application Screening, Data Enrichment etc)
• Channels → Filter agents based on where they are used (for example: Chat, SMS or Voice)
• Date Filter → Filter agents based on when they were last updated

Managing Created Agents

You can manage each agent directly from the AI Agents list.

More Options (⋯)

Click the three-dot menu next to an agent to access additional actions:

• Edit - Open the agent and update its configuration.
• Archive - Remove the agent from active use without deleting it.
• Duplicate - Create a copy of the agent with the same configuration.
• View Analytics - View performance and activity details for the agent.

Understanding AI Agent Different Statuses

Knowledge Base

The Knowledge Base section is used to manage the content that AI agents use to answer candidate’s questions during conversations with the Agent.

Agents refer to this content when a candidate asks a question, so keeping it accurate and updated is important. Upon initial access, the screen is empty, presenting suggestions to initiate the creation of a new agent. Once agents are created, they will be displayed in this area.

Articles

The Articles section shows all the knowledge base articles for the selected brand. These articles are used by agents to answer candidate questions.

Articles List

The Article table displays all the available articles, where each row represents one article.

Topic

The Topic displays the name of the article, which will help you identify what the article is about. 

You can click the topic name to view or edit the article.

Language

The Language column displays the language in which the article is. This is useful to identify and manage content across multiple languages.

The Agent can respond only in the language it is configured for. Make sure all Knowledge Base content and instructions are written in the same language that you want the Agent to support. 

If you need the Agent to support a different language, translate the entire content (including instructions and FAQs) into that language.

Content

The Content column displays a short preview of the article, helping you quickly understand its contents without opening it.

Actions

Each article has the following options:

• Edit - Open and update the article.
• Delete - Remove the article from the list.

Add Article

To create a new article, click + Add Article. From here, you can add new questions and answers for your agents.

Creating an Article

Creating an article allows you to define how the agent should respond to specific candidate questions. Each article acts as a knowledge entry that the agent can use during conversations.

When a candidate asks a question, the agent looks for a matching article based on the topic and uses the content you have provided as the response.

Steps to create an article

1. Click + Add Article.

2. Enter the Topic. This is the title of the article
3. Select the Language. Choose the language for the article as per your requirement (Applicable only when the multilingual feature is enabled).
4. In the Content text box, enter the response content that the agent should use. It is recommended that you keep the content clear and relevant.
5. Click Save to save the article. The article will be added to the list after you save it.

Agents can then use this content in the agent builder to respond when a matching question is asked.

Editing an Article

Editing an article allows you to update the existing content used by the agent to respond to candidate questions. You can modify the topic, language, or content to keep responses accurate and up to date.

Steps to edit an article

You can follow the steps below to edit an article:

1. Go to the Articles section.
2. Find the article you want to edit/update.
3. Click the Edit (✏️) icon.

4. Update the Topic, Language, or Content as needed for your article.

5. Click Save to save the changes. Once saved, AI Agents use the latest version of the article in future conversations. 

Best Practices for Articles

• Use clear and simple language
• Keep the content short and to the point
• Make sure the response directly answers the question
• Avoid adding unnecessary information
• Use exact and correct information
• Update articles regularly to keep them accurate
• Avoid duplicate articles for the same topic
• Ensure the topic clearly represents the question or intent

Default Fallback

The Default Fallback is the message shown when the agent is not able to find a relevant answer from the knowledge base.

During a conversation, the agent tries to match the candidate’s question with the available articles. If no matching article is found, the agent uses the fallback message to respond instead of leaving the question unanswered.

This helps maintain a smooth conversation and ensures that the candidate always receives a response.

What this section shows

• Displays the current fallback message configured for the selected brand
• Shows the exact message that will be sent when no answer is available
• Allows you to quickly view and update the fallback response

Note: A default fallback message is already available for each of the enabled languages. The agent will use this message based on the language of the conversation. You can update these fallback messages based on your requirements.

How to update the fallback message

You can follow the steps below to update the fallback message:

1. Go to Default Fallback
2. Click on the Edit Pencil icon

3. Update the message

4. Click Save to save your changes. Once saved, the agent will use this message whenever it cannot find a relevant answer.

Best practices

• Keep the message simple and polite
• Let the candidate know the question was not understood
• Guide them to rephrase or ask a different question
• Avoid leaving the message too generic or unclear