---
title: "Architecture overview"
manual: "AI Assistant"
version: "main"
source: "Introduction/ArchitectureOverview.rst"
rendered: "2026-10-03T14:54:45+00:00"
---

# Architecture overview {#architecture-overview}

AI Assistant combines the framework-independent `madj2k/ai-core` package with
TYPO3-specific integration code.

```text
madj2k/ai-core
├── Assistant/      Runtime, context, pipelines and prompts
├── Connection/     Provider contracts and clients
└── Indexing/       Documents, adapters, chunking and vector writes

madj2k/ai-assistant
├── Assistant/      TYPO3 profiles, memory and traces
├── Backend/        Configuration and diagnostics
├── Connection/     TYPO3 connection records and registries
├── Indexing/       TYPO3 sources and persisted indexing state
├── Controller/     Frontend and backend controllers
└── Command/        CLI commands
```

## High-level data flow {#high-level-data-flow}

```mermaid
flowchart LR
    U[User] --> C[ChatController]
    C --> O[Assistant Orchestrator]
    O --> P[Pipeline]
    P --> Q[Query processing]
    P --> R[Retriever]
    R --> V[Vector store]
    P --> A[Answer generation]
    A --> L[AI connection]
    P --> T[Pipeline traces]
    A --> U
```

## Indexing flow {#indexing-flow}

```mermaid
flowchart LR
    S[Source] --> I[Indexer]
    I --> D[IndexableDocument]
    D --> C[TextChunker]
    C --> E[Embedding request]
    E --> AI[AI connection]
    C --> W[VectorDocument]
    W --> VS[Vector store connection]
    I --> State[IndexerSource state]
```

## Registry-based extension points {#registry-based-extension-points}

Most technical extension points are collected by TYPO3's dependency injection
container and exposed through registries. This keeps the core pipeline generic:
new processors, connectors, indexers and adapters can be added without changing
central orchestration code.