Schemas

Attributes to be used in the computation.

Various filter types to filter the execution result.

Metrics to be computed.

Metrics to be referenced from other AFM objects (e.g. filters) but not included in the result.

(EXPERIMENTAL) Override definitions of catalog metrics for this request. Allows substituting a catalog metric's MAQL definition without modifying the stored definition.

Object ID.

Object type, e.g. dashboard.

Workspace ID.

An analytical dashboard identifier.

Additional details to be included in the automated message.

External recipients of the automation action results.

Additional information for the automation.

A notification channel identifier.

A list of tags.

resultId to cancel token pairs

Top level executable entity. Combination of [A]ttributes, [F]ilters & [M]etrics.

Specifies how the result data will be formatted (dimensions) and which additional data shall be computed (totals).

Various settings affecting the process of AFM execution or its result

Response to AFM execution request body

List of identifiers of the attributes to get the valid descendants for.

Map of attribute identifiers to list of valid descendants identifiers.

Top level executable entity. Combination of [A]ttributes, [F]ilters & [M]etrics.

Key columns. Defaults to first inferred column.

Type of usage counter, e.g. AI_QUERIES.

Current usage value after this request.

Entitlement limit. 0 means unlimited.

Various filter types to filter execution result. For anomaly detection, exactly one dataset is specified in the condition. The AFM may contain multiple date filters for different datasets, but only the date filter matching the dataset from the condition is used for anomaly detection.

Metrics to be computed. One metric if the alert condition is evaluated to a scalar. Two metrics when they should be evaluated to each other.

Attributes to be used in the computation.

Metrics to be referenced from other AFM objects (e.g. filters) but not included in the result.

Source object type (e.g., 'dashboard', 'visualization', 'metric').

Target object type (e.g., 'visualization', 'metric', 'attribute').

If true, allows target objects that are not part of any relationship (orphans) to be included in results. If false, orphan target objects will be excluded even if they directly match the search query. Default is true (orphans are allowed).

API key for AI project - intended for frontend use.

Amplitude endpoint URL.

API key for GoodData common project - used by backend.

Optional reporting endpoint for proxying telemetry events.

Reasoning for error states

Users who created any object in the catalog

First name of the user who created any objects

Last name of the user who created any objects

User ID of the user who created any objects

List of individual CSV analysis requests.

Location of the CSV file to analyze.

CSV analysis request config.

Possible column delimiters.

Maximum number of rows to work with during header detection.

Number of rows to consider as header, if null, header will be detected.

Number of rows to return in the flight that represents analysis result. If 0, no rows are returned, if less than 0, all rows that were in the sample are returned.

List of column metadata.

Location of the analyzed file in the source data source.

Preview of the first N rows of the file.

Config used to process the CSV file.

Name of the column as specified in the file (or autogenerated one if the file has no header).

Type of the column (e.g. string, bool, etc.).

List of date formats that can be used to parse this column as date. Null if there are none.

Options for converting CSV files when reading.

Options for parsing CSV files.

Options for reading CSV files.

Reference to the date dataset to which the filter should be applied.

Date granularity for anomaly detection. Only time-based granularities are supported (HOUR, DAY, WEEK, MONTH, QUARTER, YEAR).

Sensitivity level for anomaly detection

Outlier sensitivity level.

Anomaly detection sensitivity.

Arithmetic operator. DIFFERENCE - m₁−m₂ - the difference between two metrics. CHANGE - (m₁−m₂)÷m₂ - the relative difference between two metrics.

List of attribute elements by reference

List of attribute elements by value

Header containing the information related to attributes.

Set of label values.

Format locale code like 'en-US', 'cs-CZ', etc.

ICU formatting pattern like 'y', 'dd.MM.y', etc.

Timezone for date formatting like 'America/New_York', 'Europe/Prague', etc.

Local identifier of the attribute. This can be used to reference the attribute in other parts of the execution definition.

Indicates whether to show all values of given attribute even if the data bound to those values is not available.

A value of the current attribute label.

A value of the primary attribute label.

User description, which will be visible in application.

Email - used as lookup (must be unique). For PUT method, it must be same as in URL

Field, which should be stored in metadata in authenticationId field. In PUT and POST method it must be not present, or equal to value calculated by backend (e.g. returned from previous GET).

User password. It is not returned by GET method.

Date granularity for the interval of ONCE_PER_INTERVAL trigger. Supported granularities: DAY, WEEK, MONTH, QUARTER, YEAR.

Trigger behavior for the alert. ALWAYS - alert is triggered every time the condition is met. ONCE - alert is triggered only once when the condition is met. ONCE_PER_INTERVAL - alert is triggered when the condition is met, then suppressed for the interval. If no interval is specified, it behaves as ALWAYS.

Export request object describing the export properties for dashboard tabular exports (v2 with dashboardId).

E-mail address to send notifications from.

Export request object describing the export properties and metadata for image exports.

Export request object describing the export properties and overrides for raw exports.

Cron expression defining the schedule of the automation. The format is SECOND MINUTE HOUR DAY-OF-MONTH MONTH DAY-OF-WEEK (YEAR). The example expression signifies an action every 30 minutes from 9:00 to 17:00 on workdays.

Timezone in which the schedule is defined.

Human-readable description of the cron expression.

Timestamp of the first scheduled action. If not provided default to the next scheduled time.

Export request object describing the export properties and metadata for slides exports.

Export request object describing the export properties and overrides for tabular exports.

Export request object describing the export properties and metadata for dashboard PDF exports.

List of user groups

List of users

Authentication type.

AWS Access Key ID.

AWS Secret Access Key.

AWS Session Token (for temporary credentials).

Authentication configuration.

AWS region for Bedrock.

Provider type.

Authentication type.

Azure API key.

Authentication configuration.

Azure OpenAI endpoint URL.

Provider type.

Date granularity specifying particular date attribute in given dimension.

Start of the filtering interval. Specified by number of periods (with respect to given granularity). Typically negative (historical time interval like -2 for '2 days/weeks, ... ago'). If null, then start of the range is unbounded.

End of the filtering interval. Specified by number of periods (with respect to given granularity). Value 'O' is representing current time-interval (current day, week, ...). If null, then end of the range is unbounded.

Start timestamp of the removal interval.

Bytes removed during this interval.

End timestamp of the removal interval.

Data about the whole organization's cache usage.

Map of data about the cache usage of the individual workspaces.

The analyzed time period

Attributes to analyze for significant changes

Optional filters to apply

Metric is a quantity that is calculated from the data.

The title of the measure being analyzed

The reference time period

Whether to use smart attribute selection

The analyzed time period (e.g., '2025-02')

Metric is a quantity that is calculated from the data.

The reference time period (e.g., '2025-01')

Attributes to analyze for significant changes. If empty, valid attributes will be automatically discovered.

Auxiliary measures

Exclude attributes with any of these tags. This filter applies to both auto-discovered and explicitly provided attributes.

Optional filters to apply.

Only include attributes with at least one of these tags. If empty, no inclusion filter is applied. This filter applies to both auto-discovered and explicitly provided attributes.

Whether to use smart attribute selection (LLM-based) instead of discovering all valid attributes. If true, GenAI will intelligently select the most relevant attributes for change analysis. If false or not set, all valid attributes will be discovered using Calcique. Smart attribute selection applies only when no attributes are provided.

Links to the execution result.

The change analysis result data containing significant changes.

Chat History interaction ID. Unique ID for each interaction.

Has the interaction already finished? Can be used for polling when interaction is in progress.

User question

Question -> Use Case routing. May contain final answer is a special use case is not required.

Change analysis specification.

Visualization definitions created by AI.

Error response in anything fails.

List of objects found by similarity search and post-processed by LLM.

Reasoning wrapper containing steps taken during request handling.

Text response for general questions.

Chat History thread suffix appended to ID generated by backend. Enables more chat windows.

User feedback.

Return chat history records only after this interaction ID. If empty, complete chat history is returned.

User feedback.

Response state indicating the outcome of the AI interaction.

Created and saved visualization IDs.

Chat History thread suffix appended to ID generated by backend. Enables more chat windows.

User feedback.

User text feedback for the interaction.

List of chat history interactions.

The conversation thread ID.

User question

Filter relationships and search results based on allowed relationship type combinations. When specified, only relationships matching the allowed types are returned (e.g. for view-only users).

If true, includes hidden objects in search and visualization building. If false (default), excludes objects where isHidden=true.

Maximum number of created results.

Maximum number of relevant objects included into context for LLM (for each object type).

Maximum number of search results.

List of object types to filter the search and visualization building. If empty or null, all object types are considered.

Score, above which we return found objects. Below this score objects are not relevant.

Score, above which we return found object(s) and don't call LLM to create new objects.

Chat History thread suffix appended to ID generated by backend. Enables more chat windows.

Temporary for experiments. Ratio of title score to descriptor score.

User context with ambient UI state (view) and explicitly referenced objects.

Change analysis specification.

Chat History interaction ID. Unique ID for each interaction.

Visualization definitions created by AI.

Error response in anything fails.

List of objects found by similarity search and post-processed by LLM.

Reasoning wrapper containing steps taken during request handling.

Question -> Use Case routing. May contain final answer is a special use case is not required.

Text response for general questions.

Chat History thread suffix appended to ID generated by backend. Enables more chat windows.

Tool call events emitted during the agentic loop (only present when GEN_AI_YIELD_TOOL_CALL_EVENTS is enabled).

AI usage metadata returned after the interaction (e.g. current query count vs. entitlement limit).

Number of interactions in the time window

Maximum number of interactions in the time window any user can do in the workspace

Time window in hours

Number of clusters to create

Clustering algorithm threshold

Number of clusters to create

Threshold used for algorithm

Column name

SQL column type (e.g. VARCHAR(200), BIGINT, DOUBLE)

Column name.

Specifies the attribute's column to which this label is associated.

Label type for the target attribute.

Logical Data Model type for the column.

Columns to partition by.

Warning message related to the column.

Column name.

Operand of the alert condition.

Ordered list of conversation items.

Conversation item content payload.

Parent conversation identifier.

Item creation timestamp (ISO-8601 UTC).

Conversation item identifier.

Zero-based item order within the response turn.

Author role of the item.

Optional parent item identifier.

Response/turn identifier grouping related items.

Task identifier grouping related work items.

Conversation identifier.

Conversation creation timestamp (ISO-8601 UTC).

Last activity timestamp (ISO-8601 UTC).

Organization identifier.

Owner user identifier.

Workspace identifier.

Conversation responses.

Response creation timestamp (ISO-8601 UTC).

Response identifier.

Response update timestamp (ISO-8601 UTC).

Optional user feedback.

Location of the file in the staging area to convert.

Location of the converted GeoParquet file in the staging area.

S3 path prefix to the parquet files (e.g. 'my-dataset/year=2024/'). All parquet files must be at a uniform depth under the prefix — either all directly under the prefix, or all under a consistent Hive partition hierarchy (e.g. year=2024/month=01/). Mixed layouts (files at multiple depths) are not supported.

Name of the pre-configured S3/MinIO ObjectStorage source

Name of the OLAP table to create. Must match ^[a-z][a-z0-9_]{0,62}$

Override inferred column types. Maps column names to SQL type strings (e.g. {"year": "INT", "event_date": "DATE"}). Applied after parquet schema inference.

Distribution configuration for the OLAP table.

Key configuration for the table data model.

Cap VARCHAR(N) to this length when N exceeds it. 0 = no cap.

Partition configuration for the table.

CREATE TABLE PROPERTIES key-value pairs. Defaults to {"replication_num": "1"}.

List of attributes representing the dimensionality of the new visualization

List of filters to be applied to the new visualization

Proposed ID of the new visualization

List of metrics to be used in the new visualization

Suggestions for next steps

Proposed title of the new visualization

Visualization type requested in question

Visualization config for smart-function rendering.

Saved visualization ID.

List of created visualization objects

DEPRECATED: Use top-level reasoning.steps instead. Reasoning from LLM. Description of how and why the answer was generated.

List of suggestions for next steps. Filled when no visualization was created, suggests alternatives.

Whether to try to automatically dict-encode string / binary data.

The maximum dictionary cardinality for autoDictEncode.

Whether to check UTF8 validity of string columns.

Information about the column types in the table.

The character used as decimal point in floating-point and decimal data.

Sequence of strings that denote false Booleans in the data.

The names of columns to include in the Table. If empty, the Table will include all columns from the CSV file. If not empty, only these columns will be included, in this order.

If false, columns in includeColumns but not in the CSV file will error out.

Sequence of strings that denote nulls in the data.

Whether quoted values can be null.

Whether string / binary columns can have null values.

Sequence of strptime()-compatible format strings, tried in order when attempting to infer or convert timestamp values.

Sequence of strings that denote true Booleans in the data.

The column name.

Whether the data in the given column can be null.

The column type.

Map of column names to date formats to use when parsing them as dates.

Options for converting CSV files when reading.

Options for parsing CSV files.

Options for reading CSV files.

Method used to read the CSV file.

The character delimiting individual cells in the CSV data.

Whether two quotes in a quoted CSV value denote a single quote in the data.

The character used optionally for escaping special characters or false to disable escaping.

Whether empty lines are ignored in CSV input.

Whether newline characters are allowed in CSV values.

The character used optionally for quoting CSV values or false to disable quoting.

Whether to autogenerate column names if columnNames is empty.

How many bytes to process at a time from the input stream.

The column names of the target table.

The character encoding of the CSV data.

The number of rows to skip before the column names (if any) and the CSV data.

The number of rows to skip after the column names.

Whether to use multiple threads to accelerate reading.

Override value.

Format override.

Metric title override.

Map of CustomLabels with keys used as placeholders in document.

Map of CustomMetrics with keys used as placeholders in document.

Dashboard object ID.

Widgets currently visible on the dashboard.

If true, the export will contain the information about the export – exported date, dashboard filters, etc.

Merge equal headers in neighbouring cells. Used for [XLSX] format only.

Set page orientation. (PDF)

Set page size. (PDF)

List of rules

List of user groups

List of users

Filename of downloaded file without extension.

Requested tabular export type.

List of filters that will be used instead of the default dashboard filters.

Map of tab-specific filter overrides. Key is tabId, value is list of filters for that tab.

Additional settings.

List of widget identifiers to be exported. Note that only one widget is currently supported.

Dashboard identifier

Filename of downloaded file without extension.

Requested tabular export type.

List of filters that will be used instead of the default dashboard filters.

Map of tab-specific filter overrides. Key is tabId, value is list of filters for that tab.

Additional settings.

List of widget identifiers to be exported. Note that only one widget is currently supported.

Mapping from dimension items (either 'localIdentifier' from 'AttributeItem', or "measureGroup") to their respective values. This effectively specifies the path (location) of the data column used for sorting. Therefore values for all dimension items must be specified.

Mapping from dimensions to data column locators.

Parameter name.

Parameter value.

Identifier of a user or user-group.

Data source ID.

ID of table.

Data source entity type.

Path to table.

Id of the AI Lake Database instance

Name of the AI Lake Database instance