Knowledge Base Building with AI

Convert the following notes into a knowledge base article.

Topic: [ARTICLE TOPIC — e.g., "how to request access to the production database"]

Audience: [WHO READS THIS — e.g., "new engineers in their first 2 weeks"]

Raw notes:

[PASTE YOUR NOTES]

Structure the article as:

## [TITLE]

**Last updated:** [TODAY'S DATE]

**Owner:** [LEAVE BLANK — I WILL FILL IN]

### Summary (2 sentences: what this article covers and when someone needs it)

### Prerequisites (what the reader needs before following this)

### Steps (numbered, specific)

### Common problems and fixes

### Related articles (leave as placeholders: [LINK TO: topic])

Highlight any information that is missing from the notes and needs to be filled in before publishing.

Create a reusable knowledge base template for the following type of article.

Article type: [DESCRIBE — e.g., "troubleshooting guides for infrastructure failures"]

Team: [TEAM AND CONTEXT — e.g., "SRE team; engineers use this during on-call incidents"]

KB tool: [TOOL — e.g., Confluence / Notion / GitBook / Markdown in Git]

The template should include:

- Section headings with [PLACEHOLDER] labels explaining what goes in each section

- An example entry for the most important 2 sections to show the right level of detail

- A metadata block at the top: title / owner / last reviewed / applies-to

- A "when to use this article" field so readers can self-select correctly

After the template, list 3 ways this template could fail in practice (e.g., authors skip sections, placeholders not replaced).

Audit the following knowledge base for gaps and quality issues.

Team/domain: [e.g., "backend engineering onboarding for a Node.js/PostgreSQL product team"]

Existing articles (paste list of titles or paste article content):

[PASTE ARTICLE TITLES OR CONTENT]

Identify:

1. Topics a new team member would need that are not covered

2. Articles with titles too vague to be findable (suggest better titles)

3. Topics that are probably covered in multiple articles and should be consolidated

4. Articles that have likely gone stale (based on topic + typical rate of change)

Priority the gaps: which missing article would cause the most friction for a new team member in week 1?

Review this knowledge base article for staleness. The article was last updated on [DATE]. Today is [TODAY'S DATE].

[PASTE ARTICLE CONTENT]

Identify:

1. Specific claims or steps that are most likely to have changed since [DATE]

2. Tool names, version numbers, or service names that should be verified

3. Links or references that should be checked

4. Sections where the process has likely been replaced by a better approach

Rate the overall staleness risk: Low (safe for 6+ months) / Medium (review in 3 months) / High (review immediately).