Query builder for dataset properties

The query builder is a visual tool for finding datasets by their properties without writing queries by hand. Instead of memorizing query syntax, you assemble structured queries by combining conditions with AND/OR logic, and reordering them by dragging. As you build, the matching datasets update in the results list.

---

When to Use the Query Builder

Reach for the query builder whenever you'd otherwise scan dataset names or guess at tags to find what you need:

• Find datasets by specific property criteria — for example, all datasets where machine equals XYZ-100 and preprocessing_version is at least 3.0.

• Build complex queries visually — combine multiple conditions with AND/OR logic without writing or remembering query syntax.

• Try different filter combinations quickly — drag conditions into a new order to compare results on the fly.

• Save, share, or reuse a query — export it as a JSON document to drop into API calls and scripts, or hand it to a teammate.

---

How It Works

Open the query builder from the dataset browser. Each query is made of one or more conditions, which you combine with AND/OR logic.

To add a condition:

1. Enter the property key you want to use in the query.

2. Choose an operator — see the supported operators below.

3. Enter a value to match against.

Add as many conditions as you need and combine them with AND (all conditions must match) or OR (any condition can match). Drag and drop conditions to reorder them; the results list updates as the query changes.

Supported Operators

Operator	Matches when the property value…
is	is equal to the entered value
is not	is not equal to the entered value
contains	contains the entered value
does not contain	does not contain the entered value
exists	is present on the dataset, regardless of value
does not exist	is not present on the dataset, regardless of value
Greater than (>)	is numerically greater than the entered value
Greater than or equal (>=)	is numerically greater than or equal to the entered value
Less than (<)	is numerically less than the entered value
Less than or equal (<=)	is numerically less than or equal the entered value

Exporting a Query

When your query is ready, export it as a JSON document to reuse in API integrations or to share with teammates. The export captures the full query — every condition, its operator, and the AND/OR grouping that ties them together.

Below you can find an example how the the query shown above looks as JSON.

{
  "and": [
    [
      "=",
      "domain",
      "manufacturing"
    ],
    [
      "exists",
      "material",
      true
    ],
    {
      "or": [
        [
          "=",
          "project_owner",
          "[e-mail]"
        ],
        [
          "=",
          "project_owner",
          "[e-mail]"
        ]
      ]
    }
  ]
}

💡 The structure above is illustrative — confirm the exact field names against what the builder actually exports before publishing.

---

Current Limitations

💡 The query builder currently supports dataset properties. Support for additional property types, such as those for dataset version properties, will be added in future releases.

---

Related Pages

• Dataset Properties — Structured key-value metadata on datasets and dataset versions

• Custom Properties — Properties on individual datums (files)

• Organize Files with Tags — How to label datums with tags