Knowledge Agents & AI answers

What are Knowledge Agents?

Knowledge Agents are customizable AI-powered assistants trained to deliver fast, relevant answers by using custom prompts and leveraging Guru’s Knowledge Graph. These agents act as strategic partners, helping any team like IT, HR, and Sales scale their knowledge-sharing efforts by connecting to key systems and sources. With Knowledge Agents working seamlessly in the background, your team can focus on what matters, while getting the right information wherever they’re working.

How do Knowledge Agents work?

Knowledge Agents are like mini custom-trained AI models within your larger Guru Answers setup. When creating one, admins can give Agents a unique name and icon, decide which data sources it can search, and give specific teams or users permission to use the Agent. The Agent can be trained to give more personalized responses by creating prompts that control things like tone, length, and technical detail in order to give more relevant answers.

Where can I use a Knowledge Agent?

Knowledge Agents can be used in the Guru web app, Google Chrome browser extension, and Slack. For example, an admin can enable an IT-specific Knowledge Agent within an IT support Slack channel, so all questions in that channel will be proactively answered using that specific Agent, producing the most relevant answer.

Users can easily switch between different Agents in the web app, Slack, or via API to get the most relevant answer based on the nature of the question they’re asking. Knowledge Agents can also be linked to specific web pages, so they automatically chime in when someone visits those URLs.

If your team is using Guru’s Custom Pages features, Knowledge Agents can also be assigned to specific Pages so that questions are automatically answered using the assigned Agent on the Page.

Why should my team create a Knowledge Agent?

Imagine someone asks in Guru, “Does our dental insurance cover braces?” Without a Knowledge Agent, Guru might pull from a variety of outdated or irrelevant sources, including material from other departments. This could result in incomplete or incorrect answers that lack context.


Now, picture an HR manager creating a Knowledge Agent specifically for employee benefits. This Agent is trained to deliver only HR-approved information, ensuring the response is accurate and reliable. With the Knowledge Graph working in the background, the Agent understands the context of the question and the person asking—such as the employee's role, location, or specific benefit plan—allowing it to pull the most relevant and personalized answer from trusted sources. This ensures faster, more accurate responses that match the needs of the person asking.