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

# KB Enrichment setup

> Deployment and configuration of the AI Platform KB Enrichment service, which provides design-time document intelligence over knowledge-base documents.

<Info>
  **Available starting with FlowX.AI 5.9.1.** `ai-platform-kb-enrichment` replaces the decommissioned `di-platform` service.
</Info>

## Overview

`ai-platform-kb-enrichment` provides **design-time document intelligence** over knowledge-base documents — chunking, enrichment, and preparing content for indexing. It is a Python service in the AI Platform, exposing gRPC and consuming a Kafka enrichment-request topic.

It deploys as a sub-chart of the AI Platform umbrella Helm chart like every other AI service (listening on port `9100` via `SERVICE_PORT` in Kubernetes), but has a few extra deployment requirements covered below.

## Prerequisites

<CardGroup cols={2}>
  <Card title="Infrastructure" icon="server">
    * AI Platform chart deployed (KB Enrichment is a subchart)
    * PostgreSQL instance (shared with the other AI Platform databases)
    * S3-compatible object storage (MinIO, AWS S3, …)
    * Kafka
  </Card>

  <Card title="Platform dependencies" icon="puzzle-piece">
    * `doc-parser` deployed (`doc-parser.installed: true`)
    * An organization `TEXT_GENERATION` LLM capability configured
    * Egress to `huggingface.co` on first start (or a pre-seeded model cache)
  </Card>
</CardGroup>

## Deployment

KB Enrichment is enabled as part of the AI Platform chart. The dependencies below are provisioned automatically by the FlowX Helm charts in standard deployments — the notes call out what each one needs.

### PostgreSQL database

KB Enrichment uses a dedicated **`kbenrichment`** database on the same instance and credentials as the other AI Platform databases (for example, next to the `embedder` database). The FlowX Helm dependencies chart provisions it automatically.

<Info>
  KB Enrichment uses PostgreSQL only for service/job state — vector embeddings are stored in Qdrant. The `knowledgebases_design` Qdrant collection is auto-created by the `embedder` service; no provisioning step is required.
</Info>

### Object storage

Parse offloads go to the **`kb-enrichment-bucket`** bucket (auto-created on MinIO), shared with Knowledge Base Indexer v2: the indexer writes offload documents and KB Enrichment reads them back.

| Environment Variable           | Description                                                                           | Default Value          |
| ------------------------------ | ------------------------------------------------------------------------------------- | ---------------------- |
| `KB_ENRICHMENT_OFFLOAD_BUCKET` | Optional override for the offload bucket name. Leave unset to use the shared default. | `kb-enrichment-bucket` |

### Model cache and egress

On first start the service downloads its `model2vec` chunking model from `huggingface.co`. The Helm chart already points the model caches at writable paths, so no manual configuration is needed:

| Environment Variable | Description                  | Default Value |
| -------------------- | ---------------------------- | ------------- |
| `HF_HOME`            | Hugging Face cache directory | `/tmp`        |
| `XDG_CACHE_HOME`     | XDG cache directory          | `/tmp`        |
| `NLTK_DATA`          | NLTK data directory          | `/tmp`        |

<Warning>
  Air-gapped environments must allow the `huggingface.co` download on first start or pre-seed the model cache; otherwise the service cannot initialize its chunking model.
</Warning>

### Kafka topic

KB Enrichment consumes enrichment requests produced by Knowledge Base Indexer v2:

| Topic                                                               | Partitions | Direction                            |
| ------------------------------------------------------------------- | ---------- | ------------------------------------ |
| `ai.flowx.ai-platform.knowledgebase.internal.enrichment-request.v1` | 3          | Indexer v2 → KB Enrichment (consume) |

The topic is provisioned automatically by the FlowX Kafka chart — no manual creation is required for standard deployments.

## Required platform configuration

<Steps>
  <Step title="Turn on doc-parser">
    AI Base ships with `doc-parser.installed: false`. Document ingestion through KB Enrichment requires `doc-parser` to be deployed — set `doc-parser.installed: true` in your AI Platform values. See the [Document Parser setup guide](/5.9/setup-guides/document-parser-setup).
  </Step>

  <Step title="Seed the LLM configuration">
    The organization needs a `TEXT_GENERATION` capability with a default model and provider API key configured in the **AI Providers** UI. Without it, enrichment fails with a `no capability TEXT_GENERATION` error. See [AI providers](/5.9/ai-platform/ai-providers).
  </Step>
</Steps>

<Info>
  Service accounts and Keycloak clients follow the standard AI Platform pattern — nothing new is required for KB Enrichment.
</Info>

## Verification

<Steps>
  <Step title="Check the pod">
    ```bash theme={"system"}
    kubectl get pods -l app=kb-enrichment   # 5.9.1: -l app=ai-platform-kb-enrichment
    ```
  </Step>

  <Step title="Confirm dependencies">
    Verify the `kbenrichment` database exists, the `kb-enrichment-bucket` was created on object storage, and the enrichment-request topic is present.
  </Step>

  <Step title="Review logs">
    On a healthy first start the logs show the `model2vec` model downloading and the service connecting to PostgreSQL and Kafka. A `no capability TEXT_GENERATION` error means the organization LLM configuration is missing (see above).
  </Step>
</Steps>

## Related resources

<CardGroup cols={2}>
  <Card title="AI Platform setup" icon="robot" href="/5.9/setup-guides/ai-platform-setup">
    Full AI Platform deployment and shared configuration
  </Card>

  <Card title="Document Parser setup" icon="file-lines" href="/5.9/setup-guides/document-parser-setup">
    Required dependency for document ingestion
  </Card>

  <Card title="AI providers" icon="plug" href="/5.9/ai-platform/ai-providers">
    Configure the TEXT\_GENERATION LLM capability
  </Card>

  <Card title="Qdrant setup" icon="database" href="/5.9/setup-guides/qdrant">
    Vector store for knowledge-base embeddings
  </Card>
</CardGroup>


## Related topics

- [AI Platform setup](/5.9/setup-guides/ai-platform-setup.md)
- [Deployment guidelines v5.10](/release-notes/v5.x/v5.10.0-july-2026/deployment-guidelines-v5.10.md)
- [FlowX.AI architecture](/5.9/docs/platform-overview/flowx-architecture.md)
- [Deployment guidelines v5.9.2](/release-notes/v5.x/v5.9.2-july-2026/deployment-guidelines-v5.9.2.md)
- [Deployment guidelines v5.9.1](/release-notes/v5.x/v5.9.1-june-2026/deployment-guidelines-v5.9.1.md)