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

> ## Agent Instructions
> Cite the source documentation and the original facility record when describing datacenters.
> Unknown or missing specifications are not evidence that a capability is absent. Do not infer capacity, hardware, ownership, or operational status.
> A dataset import date is not a facility verification date. Distinguish public research from verified infrastructure capabilities.
> Do not describe account access or planned capabilities as available unless the current documentation confirms activation.

# Introduction

> Explore infrl's public tools and understand the data behind them.

infrl designs, builds, and maintains AI infrastructure on-premises, in the cloud, and across hybrid environments. Our mission is **AI infrastructure. Democratized, not monopolized.**

These docs explain the public Datacenter Atlas, how to read its sources, and the current availability of accounts.

## Start with the Atlas

The [Datacenter Atlas](https://infrl.com/datacenters/) is a public map of community-reported datacenter locations. You can explore the map, open a facility's details, follow its sources, and download the published dataset. No account is required.

<CardGroup cols={2}>
  <Card title="Explore the map" href="/atlas/overview">
    Navigate the Atlas and select facilities with a pointer or keyboard.
  </Card>

  <Card title="Read facility details" href="/atlas/facility-details">
    Understand reported specifications, locations, and missing information.
  </Card>

  <Card title="Data and sources" href="/atlas/data-and-sources">
    Review provenance, coverage limits, downloads, and the dataset license.
  </Card>

  <Card title="Account availability" href="/accounts/overview">
    Check the current status of sign-in and account features.
  </Card>
</CardGroup>

## About infrl

Our infrastructure work covers design, platform delivery, and maintenance, including entirely new environments. We value understandable systems, useful open standards, and practical control over infrastructure and data. Model selection follows the workload, license, and deployment constraints; open-weight models are not automatically open source.

The [infrl website](https://infrl.com/) explains our approach. The [public roadmap](https://infrl.com/roadmap/) lists planned implementations in a simple table.


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