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

# Data Sources

> The raw data you bring into a room — structured and unstructured, from files to live APIs.

A **Data Source** is the raw data you bring into a [room](/concepts/rooms). Data sources are the foundation of everything else in Corvic — workflows process them, Corvic Tables compose them, and agents reason over them.

## Structured vs. unstructured

Corvic works natively across both kinds of data:

| Type             | What it is                                            | Examples                                      |
| ---------------- | ----------------------------------------------------- | --------------------------------------------- |
| **Structured**   | Tabular data with a consistent schema and primary key | Parquet, CSV, DSV, TSV                        |
| **Unstructured** | Information that doesn't fit a tabular model          | Documents, spreadsheets, images, audio, video |

<Note>
  For the complete list of supported formats by category, see [Data Types](/features/data-types).
</Note>

## Ways to bring in data

There are several ways to get data into a room:

* **File upload** — upload files directly from your computer.
* **Data connectors** — connect live to Amazon S3, Azure Blob Storage, Google Cloud Storage, or Snowflake, with optional sync schedules.
* **Web sources** — search the web in plain language and extract structured tables.
* **API fetch** — pull data from any external API using a Python action and the API Secret Vault.

<Card title="Data Sources (Feature Guide)" icon="database" href="/features/data-sources">
  Step-by-step instructions for each method.
</Card>

## How data is processed

Once a source is added, Corvic builds the right pipeline for it automatically — structured data becomes tabular Corvic Tables, while documents, images, audio, and video are routed through the appropriate extraction engine (OCR/parsing, transcription, vision, and more). You then transform and embed that data through [workflows](/features/workflows).

## Related Concepts

<CardGroup cols={2}>
  <Card title="Data Types" icon="shapes" href="/features/data-types">
    Every supported file format.
  </Card>

  <Card title="Corvic Tables" icon="eye" href="/concepts/feature-views">
    Compose data sources into entities.
  </Card>

  <Card title="Rooms" icon="folder" href="/concepts/rooms">
    Where data sources live.
  </Card>

  <Card title="Workflows" icon="window" href="/features/workflows">
    Process and embed your data.
  </Card>
</CardGroup>