> For the complete documentation index, see [llms.txt](https://docs.cherryai.com.cn/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.cherryai.com.cn/docs/en-us/knowledge-base/cases.md).

# Knowledge Base Use Cases

Whether a knowledge base is reliable does not depend on the amount of material, but on whether the boundaries are clear, whether the sources are maintainable, and whether real questions can consistently retrieve the correct evidence.

{% hint style="info" %}
The parameters below are only a starting point. First run import, retrieval, and usage with 3–10 representative materials, then expand based on a fixed question set.
{% endhint %}

## Design first using the same method

{% stepper %}
{% step %}

### 1. Clearly define the final task

Clarify the judgment or deliverable the user needs to make, such as checking a policy, troubleshooting a fault, or generating a research report.
{% endstep %}

{% step %}

### 2. Define the material boundaries

Only materials that should be retrieved together during use belong in the same knowledge base. Content with different permissions, lifecycles, product models, or versions should be separated first.
{% endstep %}

{% step %}

### 3. Choose the sources and update method

Specify who maintains documents, notes, directories, and web pages, when they are replaced, and whether historical versions need to be retained.
{% endstep %}

{% step %}

### 4. Prepare acceptance questions

Prepare 3–10 real questions covering exact facts, conditional rules, colloquial phrasing, and easily confused versions.
{% endstep %}

{% step %}

### 5. Adjust the retrieval plan

Keep the configuration simple at first. Add embeddings when BM25 is not enough to handle synonyms; add reranking when the correct candidates appear but the order is unstable.
{% endstep %}

{% step %}

### 6. Connect to a dialogue or Agent

Use ordinary conversation for one-shot Q\&A; bind an Agent when multi-step research, comparison, and file delivery are needed. Verify sources item by item before going live.
{% endstep %}
{% endstepper %}

## User case 1: Employee policy Q\&A

### Goal

Enable employees to look up travel approval, lodging standards, and reimbursement exceptions, and open the sources to verify the original text.

### Material organization

* Knowledge base: \[Employee Travel Policy]
* Entry: \[Travel Approval Process]
* Entry: \[Lodging Standards Quick Reference]
* Entry: \[Travel FAQ]

<figure><img src="https://1658303467-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F0Ut5BptC3t8CtSU1UWpM%2Fuploads%2FoYCk8ddQNgaZxNQ65As2%2Fclipboard.png?alt=media&amp;token=6a24e9c5-3456-492b-bf99-fc7cd426cf30" alt="由多条制度资料组成并全部就绪的员工差旅制度知识库"><figcaption><p>Policy clauses and FAQs are maintained separately; when one is updated, the entire set does not need to be redone.</p></figcaption></figure>

### Recommended configuration

| Item                | Starting point                                  | When to adjust                                                            |
| ------------------- | ----------------------------------------------- | ------------------------------------------------------------------------- |
| Retrieval           | Start with BM25                                 | Add embeddings when employee phrasing differs greatly from policy wording |
| Reranking           | Do not use for now                              | Enable when the correct candidate appears but the order is unstable       |
| Material version    | Keep only the current version                   | When historical audits require coexistence, include the year in the name  |
| Answer requirements | Separate the conclusion, conditions, and source | Clearly mark when the material does not specify                           |

### Acceptance questions

1. What is the maximum hotel cost for a business trip?
2. Can overseas car rentals be reimbursed?
3. Who approves when the estimated total cost exceeds 5,000 yuan?
4. How is it handled if a hotel is booked without approval?

### Dialogue prompt

> Answer only based on the "Employee Travel Policy." Give the conclusion first, then list the applicable conditions and sources; if the material does not specify, write "The policy does not specify" and do not fill in with common sense.

{% hint style="success" %}
When acceptance passes, the same policy can be hit by both the original wording and colloquial wording of the question, and the amounts, roles, and conditions in the answer can be directly supported by citations.
{% endhint %}

## User case 2: Product after-sales assistant

### Goal

Organize the official manual, fault codes, and reviewed cases so customer service can first provide safe, traceable troubleshooting suggestions.

### Material boundaries

| Knowledge base or material group | Content                                                  | Maintenance principle                                 |
| -------------------------------- | -------------------------------------------------------- | ----------------------------------------------------- |
| Official manual                  | Specifications, warranty boundaries, standard procedures | Keep model and document version                       |
| Fault codes                      | One fault per subsection                                 | Mark applicable firmware and device model             |
| Reviewed cases                   | Cases with confirmed causes and solutions                | Unreviewed chat records must not be imported directly |

When rules differ significantly across models, split them into independent knowledge bases by model to avoid the same fault code competing with itself.

### Retrieval and Agent configuration

* For PDFs, first spot-check the table of contents, tables, and two-column body text.
* Fault codes rely on precise terminology; keep BM25.
* When customer descriptions are more colloquial, add an embedding model.
* Bind the official manual and reviewed cases to the after-sales Agent, and enable only \[Knowledge Base Search].

> Troubleshoot in three steps based on the device model, fault code, and symptoms. Indicate at each step whether the basis comes from the official manual or a reviewed case. When disassembly, electricity, or data erasure is involved, first warn of the risk and wait for confirmation.

### Acceptance criteria

* Do not apply steps from other models to the current model.
* Safety warnings appear before the operation steps.
* Official rules and case recommendations are kept separate.
* When there is no supporting material, transfer to a human agent and do not guess.

## User case 3: Research materials and reports

### Goal

Extract verifiable evidence from papers, interview notes, and web snapshots, then have an Agent generate a source-backed comparison report.

### Material organization

* Build the knowledge base around the research questions; do not stuff all papers into one large repository.
* Include author, year, and short title in the file name.
* Mark interview notes with the interviewee's role, date, and whether they may be cited.
* For web materials, record the crawl date because the knowledge base stores imported snapshots.

<figure><img src="https://1658303467-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F0Ut5BptC3t8CtSU1UWpM%2Fuploads%2FmfHDAsphgab351BnEKkY%2Fclipboard.png?alt=media&amp;token=0fc8847f-23e0-4d93-8938-26e818df4b8f" alt="从提出真实问题、检查召回、定位问题到只调整一项并重新索引复测的知识库质量闭环图"><figcaption><p>First verify source coverage in the research repository with fixed questions, then hand it to the Agent for cross-document synthesis.</p></figcaption></figure>

### Agent prompt

> Find evidence for "why users abandon first-time setup" in the bound research knowledge base. First list the original viewpoints and limitations by source, then summarize consensus, divergences, and hypotheses to be verified. Finally generate a Markdown report; do not write inferences as the interviewees' original words.

### From evidence to deliverable

<figure><img src="https://1658303467-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F0Ut5BptC3t8CtSU1UWpM%2Fuploads%2FuAux5cNPlWeEphenKT5f%2Fclipboard.png?alt=media&amp;token=4875b86d-943e-4aea-85db-9085048edafe" alt="资料来源经过知识库召回和 Agent 整理后形成文字文件或多语言图片的内容工作流图"><figcaption><p>Keep the evidence and limitations first, then let the Agent organize them into a report; do not let the finished product obscure the original sources.</p></figcaption></figure>

### Acceptance criteria

* Consensus is supported by at least two independent sources.
* Differences retain their own conditions and are not forced to merge.
* Citations, inferences, and recommendations are clearly labeled.
* Web snapshots and paper versions are traceable.

## Configuration notes: reusable design table

| Item                 | Questions to answer                                                                             |
| -------------------- | ----------------------------------------------------------------------------------------------- |
| Goal                 | What judgment or result does the user ultimately need to make or deliver?                       |
| Boundaries           | Which materials should be retrieved together, and which must be separated?                      |
| Source               | How should files, notes, directories, and web pages be updated?                                 |
| Parsing              | Which types of documents are most prone to OCR, table, or ordering issues?                      |
| Retrieval            | Is BM25 sufficient? When do embeddings and reranking become necessary?                          |
| Acceptance questions | Which 3–10 questions represent real usage?                                                      |
| Failure handling     | What should be done when there are no results, conflicting versions, or no supporting material? |
| Maintenance          | Who is responsible for replacing materials, reindexing, and backups?                            |

{% hint style="warning" %}
Do not treat "imported a lot of material" as the completion criterion. The more material there is, the more duplicate versions, mixed permissions, and noise competition need to be explicitly managed.
{% endhint %}

## FAQ

<details>

<summary>Should policies, manuals, and cases be placed in the same knowledge base?</summary>

Look at whether they should be retrieved together in the same question, and whether permissions and update cycles are consistent. When the differences are obvious, splitting the repository makes it easier to control source boundaries.

</details>

<details>

<summary>Can the case repository directly import all customer service chats?</summary>

Not recommended. First review the causes, solutions, and privacy content, and import only confirmed, reusable cases.

</details>

<details>

<summary>What needs to be done before expanding the materials?</summary>

Keep a fixed set of acceptance questions, import in batches, and retest. When new materials degrade the results, you can quickly locate which batch caused the issue.

</details>

## Continue reading

<table data-view="cards"><thead><tr><th></th><th></th><th data-hidden data-card-target data-type="content-ref"></th></tr></thead><tbody><tr><td><strong>Knowledge Base Beginner's Guide</strong></td><td>First make creation, import, retrieval, and usage work.</td><td><a href="/docs/en-us/knowledge-base/knowledge-base.md">Getting Started with Knowledge Bases</a></td></tr><tr><td><strong>Use with Agent</strong></td><td>Configure multi-step research and material permissions.</td><td><a href="/docs/en-us/knowledge-base/agent.md">Use with Agent</a></td></tr><tr><td><strong>FAQ</strong></td><td>Locate materials, retrieval, or answer problems.</td><td><a href="/docs/en-us/knowledge-base/troubleshooting.md">Frequently Asked Questions</a></td></tr></tbody></table>


---

# 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 by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.cherryai.com.cn/docs/en-us/knowledge-base/cases.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

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.