Skip to main content

Data Collections

Overview​

The Data Collections module serves to aggregate multiple data sources into a single logical unit. A data collection represents a central container of data that can subsequently be utilized in AI assistants, workflows, or analytical tools in Siesta AI.

Each collection:

  • has its own name and description,
  • contains one or more data sources,
  • allows for the management and organization of data according to its purpose.

Overview of Data Collections​

On the main screen of Data Collections, a list of all created collections is displayed in the form of a table.

Displayed columns:

  • Name – the name of the data collection
  • Description – a brief description of the collection's purpose
  • Data Sources – the number of connected data sources
  • Created – the date and time of creation
  • Actions – additional options for working with the collection

At the top of the screen, the following is available:

  • collection search,
  • the Create Collection button.

Overview of Data Collections

Creating a New Data Collection​

After clicking on Create Collection, a dialog for creating a new collection opens.

Required Fields

  • Name – a unique name for the data collection (e.g., Neurology Research Results – Big Data Collection).
  • Description – a brief description of the content and purpose of the collection.

Actions

  • Cancel – closes the dialog without saving
  • Create – creates a new data collection

Create Collection Dialog

Data Collection Detail​

When a specific collection is opened, its detailed page is displayed.

Displayed information:

  • collection name,
  • creation date,
  • overview of connected data sources.

The page includes a Add Data Source button.

Data Collection Detail

Adding a Data Source to the Collection​

Clicking on Add Data Source opens a selection of data source types.

Available Options

  • Manual Upload – manual file upload
  • Google Drive (coming soon)
  • SharePoint (coming soon)
  • Azure Storage (coming soon)
  • Jira (coming soon)

In the current version, manual file upload is available.

Select Data Source Type

Data Source Configuration (Manual Upload)​

After selecting Manual Upload, a configuration form is displayed.

Configuration Fields

  • Name – the name of the data source (e.g., Big Data).
  • Description – an optional description of the content of the data source.
  • Upload Files – the option to drag files into the designated area or click to select files from the computer.
  • JSON Functions (optional) – used to define custom functions for working with data.
  • JSON Metadata Definitions (optional) – allows adding structured metadata to the data source.

Actions

  • Cancel – exits the configuration without saving
  • Confirm – saves the data source and starts its processing

Manual Upload Configuration

Data Source Status​

Each data source has its own processing status:

  • Processing – data is being analyzed and indexed
  • Processed – the data source is ready for use

The status is visible in the data sources table in the collection detail.

Connecting Data Collections to the Assistant​

Data collections are subsequently assigned to assistants in their settings. Details can be found in the section Assistant Configuration.

Typical Uses of Data Collections​

Data collections are primarily used for:

  • organizing a larger number of files,
  • aggregating data by topic or project,
  • creating a single source of truth for AI assistants,
  • reusing the same data in different workflows,
  • scaling data work without the need for duplication.

Summary​

Data collections in Siesta AI enable clear data management and their effective use across the entire platform. Well-structured collections are the foundation for quality outputs from AI assistants and automated workflows.