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

# Native document upload information

Epiq AI can process files in their native forms, extract text and metadata, and generate PDFs for document annotations, redaction, and production exports.

### Folder structure

Epiq AI preserves the original folder structure of your native documents as part of the upload.

* Epiq AI **traverses all documents recursively** under the top-level folder.
* For each file, Epiq AI **records the folder path** in the file’s metadata.
* This preserved path can be used to understand the original organization of the documents.

### Container files

Some files act as containers that store other files or attachments. Epiq AI opens these containers and processes their contents.

#### Archive and store containers

For archive and store types such as:

* `.pst`
* `.rar`
* `.7z`
* `.gz`
* `.zip`
* `.z`
* `.bz2`
* `.alzip`

Epiq AI:

* **Extracts embedded files** from the container.
* Performs extraction **recursively**:
  * If a container includes another container (for example, a ZIP file inside a ZIP file), Epiq AI also extracts and processes the nested contents.

#### MIME-based containers

For MIME-based formats such as:

* `.eml`
* `.ics`

Epiq AI:

* **Extracts attachments** from the file.
* Processes the attachments as documents, subject to supported file types and processing rules.

### OCR and image text extraction

Epiq AI uses Optical Character Recognition (OCR) to extract text from images so that image-based content can be searched and analyzed. OCR is applied to:

* **Standalone image files.** For example, image files uploaded directly.
* **Images embedded in documents** such as:
  * PDF files
  * Word documents
  * PowerPoint presentations
  * Excel workbooks

Text detected through OCR is stored as document text and becomes available for searching and review alongside text extracted from non-image content.