> For the complete documentation index, see [llms.txt](https://docs.maiagent.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.maiagent.ai/maiagent-user-guide/en/application/km.md).

# Internal Knowledge Management

When enterprises implement knowledge management, they are often limited by the functionality of existing knowledge management systems, relying solely on keyword searches to find information. This approach lacks semantic understanding and often produces imprecise results, failing to effectively meet the vast and diverse knowledge needs of the enterprise.

Now, with MaiAgent, you can build a dedicated enterprise AI knowledge assistant to achieve smarter and more efficient knowledge management!

## Use Cases <a href="#use-cases" id="use-cases"></a>

* **Professional & Industry Knowledge Q\&A**: Journals, research reports, and internal technical documents are voluminous and constantly updated — employees need to quickly locate key information rather than reading through each document.
* **Meeting Decision Traceability**: Meeting records are scattered across various locations. When requirements and decisions change repeatedly, it is difficult to quickly verify "who decided what, in which meeting."
* **Data Visualization & Analysis**: Financial and credit analysis work requires repeatedly creating charts. Manual operations are time-consuming and produce inconsistent formats — Canvas can be used to generate interactive charts directly with natural language.
* **Document & Accounting Automation**: Reimbursement receipts and bills require manual entry into systems line by line, and must comply with internal business rules (such as department code mapping).
* **Cross-Department Access Control**: Different departments and job levels have different knowledge access scopes — the knowledge base must be able to restrict search scope based on identity.

## How It Works <a href="#how-it-works" id="how-it-works"></a>

| Step                            | Description                                                                                                                  |
| ------------------------------- | ---------------------------------------------------------------------------------------------------------------------------- |
| **1. Knowledge Sources**        | Document upload (PDF/Office/images), web crawler extraction, meeting transcripts                                             |
| **2. Knowledge Base Ingestion** | The system parses documents, chunks them, and converts them to semantic vectors, then classifies by tags and metadata        |
| **3. AI Assistant Retrieval**   | RAG-based retrieval Q\&A; Canvas for visualization when charts are needed; tool calls or database queries for real-time data |
| **4. Access Control**           | Role permissions and query metadata determine which documents each user can retrieve                                         |
| **5. Internal Users**           | Users receive responses via the admin console or connected internal channels                                                 |

The key difference from traditional keyword search is that the knowledge base first chunks documents and converts them to semantic vectors, enabling the AI assistant to understand the intent and context of questions rather than just matching literal text. Responses synthesize information from multiple sources and include citation references for verification.

## Setup Workflow <a href="#setup-workflow" id="setup-workflow"></a>

### 1. Create a Knowledge Base <a href="#step1-create-knowledge-base" id="step1-create-knowledge-base"></a>

On the <mark style="color:blue;">Knowledge Base</mark> page, create a knowledge base and complete the [basic settings](/maiagent-user-guide/en/km/km-basic-settings.md), selecting the appropriate parsing and chunking methods.

### 2. Import Knowledge Sources <a href="#step2-import-sources" id="step2-import-sources"></a>

Upload existing documents, or use the [web crawler](/maiagent-user-guide/en/km/scrape-website.md) to extract internal website content. If the knowledge source is meetings, use [meeting records](/maiagent-user-guide/en/meeting-records/meeting-records.md) to automatically transcribe and [integrate into the knowledge base](/maiagent-user-guide/en/meeting-records/download-and-export.md).

### 3. Organize Tags and Metadata <a href="#step3-tags-and-metadata" id="step3-tags-and-metadata"></a>

Add [tags and metadata](/maiagent-user-guide/en/km/tags-and-metadata.md) to documents (such as department, year, document type) — this is the foundation for subsequent permission filtering and precise retrieval.

### 4. Validate Retrieval Quality <a href="#step4-test-search" id="step4-test-search"></a>

Use [search testing](/maiagent-user-guide/en/km/test-search-result.md) to input questions that would actually be asked, verify whether the retrieved passages are correct, and decide whether to adjust chunking settings or supplement data.

### 5. Create an AI Assistant and Bind the Knowledge Base <a href="#step5-create-agent" id="step5-create-agent"></a>

[Create an AI assistant](/maiagent-user-guide/en/build/setup.md) and attach the knowledge base. If chart generation is needed, set the response mode to "Agent Mode" and enable Canvas mode.

### 6. Configure Access Permissions <a href="#step6-access-control" id="step6-access-control"></a>

Use [role permission management](/maiagent-user-guide/en/org/roles/role-permission.md) and [knowledge management permissions: query metadata](/maiagent-user-guide/en/org/querymetadata.md) to ensure users from different departments and job levels can only retrieve documents they are authorized to view.

{% hint style="warning" %}
Internal knowledge bases often contain sensitive data. Before going live, verify permission settings and test retrieval results with accounts of different roles to prevent unauthorized access.
{% endhint %}

## Application Examples <a href="#examples" id="examples"></a>

* [Professional Knowledge Management](/maiagent-user-guide/en/application/km/internalkm.md): Using maritime industry journals and reports as an example, build an internal professional knowledge Q\&A assistant.
* [Meeting Requirements Management](/maiagent-user-guide/en/application/km/meeting.md): Using software project requirement changes as an example, let AI quickly trace the decision context across meeting histories.
* [Financial Analysis Assistant](/maiagent-user-guide/en/application/km/finance.md): Use Canvas to generate interactive financial charts such as revenue and gross margin with natural language.
* [Billing Assistant](/maiagent-user-guide/en/application/km/billing-assistant.md): Recognize utility, telecom, and other reimbursement receipts, and automatically apply internal business rules such as department codes.
* [Credit Analysis Assistant](/maiagent-user-guide/en/application/km/credit.md): Integrate credit data and use Canvas to generate risk assessment charts, accelerating credit review.

The following sections demonstrate the setup process with real-world examples to help you quickly apply these capabilities to your own business scenarios.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://docs.maiagent.ai/maiagent-user-guide/en/application/km.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.