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

# Developer architecture

> A simple map of the Open Source application for contributors and operators.

Most users do not need this page. Presenton runs as one product with a web interface, an API, persistent data, configurable AI services, and a local export runtime.

```mermaid theme={null}
%%{init: {"themeVariables": {"fontSize": "17px"}, "flowchart": {"curve": "linear", "nodeSpacing": 45}}}%%
flowchart LR
  People[People and applications] --> Presenton[Presenton web app and API]
  Presenton --> Data[(Your presentations and files)]
  Presenton --> AI[Your chosen AI services]
  Presenton --> Export[PPTX and PDF export]
  classDef core fill:#F4F0FF,stroke:#6D4AFF,stroke-width:2px,color:#201547;
  class Presenton core;
```

## Main project areas

| Area                | Responsibility                                                                                      |
| ------------------- | --------------------------------------------------------------------------------------------------- |
| Next.js application | Dashboard, onboarding, generation workflow, templates, settings, and slide editor                   |
| FastAPI service     | API endpoints, provider calls, document processing, persistence, assistant actions, and task status |
| Application data    | Database, uploads, generated assets, templates, memory, and exports                                 |
| Export runtime      | Renders the saved presentation to PPTX or PDF                                                       |
| External providers  | Optional text, image, and search capabilities selected by the operator                              |

The editable presentation state is shared across generation, the canvas editor, the assistant, persistence, and export. Contributors should preserve that single source of truth when adding new slide capabilities.