> ## Documentation Index
> Fetch the complete documentation index at: https://docs.usefini.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Knowledge

> How the agent knows what to say. The end-to-end flow from raw inputs to a single curated knowledge graph that powers every customer reply.

Your AI agent's reply quality is bounded by what it knows. Knowledge is the section of the product that owns *what the agent says*: where raw content enters the system, how it's consolidated into a single source of truth, how conflicts and gaps are surfaced for review, and how it's scoped per agent so different agents have different knowledge bases.

<Tip>
  Read this page first if you're setting up an agent for the first time, then drill into each subsection.
</Tip>

## The model

Fini's knowledge model is built around one principle: **Articles is the only source of truth.** Everything else becomes candidate knowledge through the review / publish workflow that feeds it.

```mermaid theme={null}
---
title: The knowledge flow
---
flowchart LR
    subgraph inputs ["INPUTS / ORIGINS"]
        direction TB
        S["Sources<br/>URLs · files · Drive<br/>Notion · Confluence · Zendesk"]
        M["Magic Articles<br/>paste raw content"]
        P["Past conversations"]
        I["Inbox feedback"]
        B["Background AI<br/>gap & conflict detection"]
    end

    subgraph consolidation ["REVIEW / PUBLISH"]
        R["Review Queue<br/><br/>surfaces gaps<br/>and conflicts"]
    end

    subgraph truth ["SOURCE OF TRUTH"]
        A["Articles<br/>The Knowledge Graph<br/><br/>agent-scoped"]
    end

    S --> R
    M --> R
    P --> R
    I --> R
    B --> R
    R --> A
```

Three things are doing the work here:

<Steps>
  <Step title="Origins are equal">
    Sources, Magic Articles, past conversations, Inbox feedback, and background AI signals are all ways new knowledge enters Fini. None of them is the source of truth on its own. A Zendesk article in your Sources is a *candidate* for knowledge, not the canonical answer.
  </Step>

  <Step title="Review Queue is the consolidation layer">
    Most new knowledge lands here first. Review's job isn't only "approve / reject", it's to surface **gaps** (questions customers asked that no article answers) and **conflicts** (two inputs that disagree) before anything goes live. Human-created drafts can also publish directly into the **Published** tab; background AI drafts always go to Review.
  </Step>

  <Step title="Articles is the knowledge graph">
    Approved and published knowledge lives in Articles, organized in folders, scoped per agent. This is what the agent retrieves from at runtime. If it's not in Articles, the agent doesn't treat it as authoritative.
  </Step>
</Steps>

## Why Sources can't be the source of truth

<Warning>
  A help center is an artifact written for humans browsing a website. It has marketing language, outdated pages, contradictions across articles, and gaps where the answer is "ask support." Pointing an agent directly at a help center inherits all of that.
</Warning>

Articles solves this by being the *curated, deduplicated, conflict-resolved* version of everything your team knows. Sources contribute to it. They don't replace it.

## The four pages

<CardGroup cols={2}>
  <Card title="Sources" icon="database" href="/en/knowledge/sources">
    Connect links, files, Google Drive, Notion, Confluence, and Zendesk help centers as raw inputs.
  </Card>

  <Card title="Magic Articles" icon="wand-magic-sparkles" href="/en/knowledge/magic-articles">
    Paste raw content (transcripts, docs, notes) and generate articles from it — either straight to live or routed through Review.
  </Card>

  <Card title="Review Queue" icon="circle-check" href="/en/knowledge/review">
    The consolidation layer. Surfaces gaps and conflicts from every origin before anything reaches Articles.
  </Card>

  <Card title="Articles" icon="book-open" href="/en/knowledge/articles">
    The knowledge graph itself. Hosts every approved article, organized in folders, with agent-specific assignment and attribute filters.
  </Card>
</CardGroup>

## Agent-specific scoping

Articles is where agent-specific scoping happens. The top-right selector lets you toggle which folders and articles each agent can retrieve from. The same underlying knowledge graph can power very different agents:

<CardGroup cols={3}>
  <Card title="Support agent" icon="headset">
    Account · Billing · Product folders
  </Card>

  <Card title="Sales agent" icon="chart-line">
    Pricing · Comparisons · Integrations folders
  </Card>

  <Card title="German-speaking agent" icon="language">
    Only German-language folders
  </Card>
</CardGroup>

<Note>
  If two agents need different knowledge bases, scope at the Articles layer. Don't duplicate Sources or split workspaces.
</Note>

## First-day setup

If you're starting from zero:

<Steps>
  <Step title="Ingest your help center under Sources">
    Use Sitemap mode if available — it's faster, more accurate, and gives you the publisher's canonical URL list.
  </Step>

  <Step title="Run bulk Magic Articles generation">
    Generate over those Sources to draft a structured knowledge base in one pass.
  </Step>

  <Step title="Open Review Queue">
    Read the drafts, the flagged duplicates, the conflict surfaces. Edit, approve, reject.
  </Step>

  <Step title="Open Articles">
    Organize approved articles into folders that map to your customer journey. Use the top-right selector to assign folders to agents.
  </Step>

  <Step title="Test in Inbox or Playground">
    Confirm the agent retrieves the right articles before exposing it to real customers.
  </Step>
</Steps>

Each step has its own page; cross-link freely while you build.


## Related topics

- [Generate knowledge](/en/api-reference/generate-knowledge.md)
- [Bulk generate knowledge](/en/api-reference/bulk-generate-knowledge.md)
- [Check knowledge jobs](/en/api-reference/check-knowledge-jobs.md)


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.