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

# Introduction

> Build, run, and scale decentralized AI workloads on ARIS.

ARIS (Autonomous Registry and Intelligent Systems) is a decentralized compute network for verifiable AI inference. You can route prompts to distributed worker nodes through one Python SDK.

<CardGroup cols={2}>
  <Card title="Ship an app" icon="rocket" href="/quickstart">
    Send your first generation request in minutes.
  </Card>

  <Card title="Run infrastructure" icon="server" href="/run-node">
    Launch worker nodes and earn from compute.
  </Card>

  <Card title="Integrate retrieval" icon="brain" href="/agentic-rag">
    Add agentic RAG and MCP tools.
  </Card>

  <Card title="Inspect API" icon="code" href="/api-reference/generate">
    Review request, response, and error formats.
  </Card>
</CardGroup>

## How ARIS works

<Steps>
  <Step title="Authenticate">
    Your app sends an API key to the registry and receives session credentials.
  </Step>

  <Step title="Select node">
    The SDK selects a healthy compute node based on availability and policy.
  </Step>

  <Step title="Run inference">
    The node executes your prompt and returns output with usage metadata.
  </Step>

  <Step title="Settle usage">
    Credits and earnings are recorded through the registry ledger.
  </Step>
</Steps>

## Choose your path

<Tabs>
  <Tab title="Application developers">
    Start with [/quickstart](/quickstart), then continue to [/python-sdk](/python-sdk) and [/advanced-usage](/advanced-usage).
  </Tab>

  <Tab title="Node operators">
    Start with [/run-node](/run-node), then set up [/monitoring](/monitoring) and [/scaling](/scaling).
  </Tab>

  <Tab title="RAG teams">
    Start with [/agentic-rag](/agentic-rag), then review [/rag-architecture](/rag-architecture) and [/mcp-integration](/mcp-integration).
  </Tab>
</Tabs>