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

# Introduction

> Programmatically extract data out of unstructured PDFs and images

The ProSights API is hosted on `https://prism-api.prosights.co`

If you are on an enterprise distribution it will be `https://firmname.prosights.co`

## Usage

<img src="https://mintcdn.com/prosights/xKFZmPY4YZRxIuw7/images/Screenshot2025-10-23at5.43.35PM.png?fit=max&auto=format&n=xKFZmPY4YZRxIuw7&q=85&s=4481d36163c8ea3ee7c55e3f54e1e43f" alt="Screenshot2025 10 23at5 43 35PM Pn" width="919" height="655" data-path="images/Screenshot2025-10-23at5.43.35PM.png" />

The Recreate Platform works with a few key concepts. Generally, the flow is:

1. Create a Recreate Request  `POST enterprise-api/recreate/`
2. Poll status until completion `GET enterprise-api/recreate/status?recreate_id={recreate_id}`
3. \[optional] Resolve data as JSON `GET /enterprise-api/recreate/{recreate_id}/to_json`

## Outputs

Recreation outputs come as authenticated download links through the status API. These download links are authenticated for a short period of time.

```c theme={null}
data={
  'recreate_id': 'cf9b1754-8dd6-4ef1-a3b4-9615411f1363', 
  'status': 'COMPLETED', 
  'xls_download_url': 'https://storage.googleapis.com/...authenticated_url...', 
  'ppt_download_url': 'https://storage.googleapis.com/...authenticated_url...',
  ...
}
```

You can also resolve these outputs as JSON through the`to_json`API.

When using the [Python SDK](/logo/python-sdk/methods) you can resolve the JSON into Pandas dataframes.