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

# File Storage

> Upload CSV, JSON, and Parquet files and use them as Zwiron pipeline sources without a separate bucket.

Zwiron includes built-in file storage so you can upload flat files and use them as pipeline sources — without needing a separate S3 bucket or file server.

<Info>File storage is included on paid plans. Start with the [14-day free trial](https://zwiron.com/login?mode=signup), then subscribe to [Individual](/platform/billing) or [talk to us](mailto:support@zwiron.com) for Enterprise.</Info>

***

## Zwiron-managed storage

Go to **Files** in the main navigation to access Zwiron's built-in file storage.

### Supported formats

| Format      | Extension  | Notes                                                      |
| ----------- | ---------- | ---------------------------------------------------------- |
| **CSV**     | `.csv`     | Auto-detects delimiter and header row                      |
| **JSON**    | `.json`    | Array of objects or single JSON object                     |
| **JSONL**   | `.jsonl`   | One JSON object per line (recommended for large files)     |
| **Parquet** | `.parquet` | Columnar format — fastest read performance for large files |
| **Excel**   | `.xlsx`    | Reads the first sheet; header row required                 |

**Maximum file size:** 1 GB per file.

### Organizing files

* Create **folders** to organize files by project, team, or pipeline
* Upload files individually or in bulk
* Files are stored with the organization and accessible to all members

### Preview and download

* Click any file to **preview** the first N rows (schema + sample data)
* Click **Download** to export the file

***

## Using a file as a pipeline source

To use a file you've uploaded as the source for an ETL pipeline:

1. Create a new connection with connector type **File (Zwiron Storage)**
2. Select the file or folder you want to use
3. Use this connection as the source in a new job

When the job runs, Zwiron reads the file, applies any configured transforms, and writes to the destination.

***

## Remote file browser

In addition to Zwiron-managed storage, you can connect to your own cloud storage and browse it directly from Zwiron. When you create a connection using:

* **Amazon S3**
* **Google Cloud Storage**
* **Azure Blob Storage**
* **MinIO**
* **SFTP**

...the connection's detail page includes a **File Browser** — a UI to navigate your bucket or server, browse folders, preview files, and select specific files or prefixes for a pipeline.

This is useful when your files are already in cloud storage and you don't want to re-upload them.

***

## Format options

When reading a file source, you can configure:

| Option          | Description                                                     |
| --------------- | --------------------------------------------------------------- |
| **Delimiter**   | For CSV — comma, tab, pipe, etc.                                |
| **Has header**  | Whether the first row contains column names                     |
| **Skip rows**   | Skip N rows at the top (for files with metadata before headers) |
| **Compression** | gzip, zstd, snappy (auto-detected for Parquet)                  |
| **Schema**      | Auto-infer types or specify column types explicitly             |

***

## File sources and scheduling

You can schedule a pipeline that reads from a file on a recurring basis. If you replace the file with a newer version (same name), the next scheduled run will read the new file automatically.

This is a common pattern for daily data deliveries from partners or vendors — they drop a new CSV in an agreed folder each morning, and Zwiron picks it up automatically.