Ignite Procurement Export API (2.0.0)

Lists the available export tables and the columns available for filtering and projection. Classification result tables expose spend_id plus the available level_N_id / level_N_name columns from the tenant's classification tree; the schema supports levels 1 through 10, with deeper levels nullable when not used.

Table ids: Your spend, supplier and contract data are served under the fixed ids spend, suppliers and contracts, classification results under classification_business_units and classification_spend_categories, and spend-based emissions under emissions. Each other classification structure (every structure except business units, spend categories and the emission structures) is its own table with the id classification_custom__<structure id>, named after the structure. It has one row per spend line classified into the structure, with spend_id and level_1_id/level_1_name through level_10_id/level_10_name. A spend line that is not in the table is not classified in that structure. The id stays the same when the structure is renamed; the table name follows the structure name, so key on the id. A rows request that would read more than 10 GB is rejected; on large structures, request only the fields you need or use an async export. Any other table you have defined in Ignite appears with the id datatable__<table id>, where the table id is the table's UUID with hyphens replaced by underscores. Your spend, supplier and contract tables are never listed under a datatable__ id, even though they have a UUID in Ignite; use the fixed ids.

Column names: name is the stable identifier and the only value accepted in fields, filters and sort. Standard columns (for example id, amount, transaction_date, supplier_id) are part of this API's contract and are not renamed. Columns you have defined yourself appear as custom__<field id>; the field id is assigned when the column is created and never changes or gets reused, so these names are safe to hard-code. Deleting a column and creating a new one with the same label produces a new field id. label is the display name you have given the column and changes whenever you rename it, so never key on label unless you have chosen keyBy: label deliberately. The row id on spend and suppliers is the same identifier the removed /export/ endpoint returned as rowId.

Every scalar column supports is_null and is_not_null. STRING columns additionally support literal substring search with contains and blank-value checks with is_blank / is_not_blank. Columns whose dataType is STRING_ARRAY (repeated string values) support only the array operators contains_any, contains_all, is_empty, and is_not_empty. Array columns are not part of the default response fields, so request them explicitly via fields to include their values in a rows response.

updated_at semantics: The updated_at column on suppliers and contracts reflects when the core record was last modified in Ignite. It does not advance when enrichment data is refreshed or when custom fields are edited. Consumers relying on updated_at for incremental sync will miss those changes; use the async export endpoint (POST /tables/{entity}/exports) for a complete point-in-time snapshot instead.

emissions columns: One row per spend transaction. emission_factor_category is the EXIOBASE product category and emission_factor_region the EXIOBASE region assigned to the transaction by classification rules, by default from the supplier's NACE code and country. Together with the transaction year they select the spend-based emission factor. emissions_kg_co2e is the transaction's spend multiplied by factor, with the spend valued in the currency given in unit (for example kgCO2e/NOK). emission_factor_source is spend_factor when a factor matched and customer_average when your average factor was used for an unmatched Scope 3 transaction, in which case customer_average_factor holds the value. emissions_kg_co2e is null when no factor applies, for example when the transaction has no category. ghgp_scope and ghgp_category give the GHG Protocol scope and category. Rows with excluded = true have 0 emissions, and excluded_reason says why: ACTIVITY_OVERRIDE (covered by activity-based data), EXEMPT (category in the Exempt group, such as taxes or internal transactions) or NON_SCOPE_3 (Scope 1 or 2, where spend-based factors do not apply). Filter excluded = false for emission totals.

Fetches scalar rows from an export table using optional field projection, typed filters, and opaque cursor pagination. Use GET /tables/ first to discover valid table ids, column names, data types, and supported filter operators. Classification result rows are identified by spend_id and contain the available level_N_id / level_N_name columns, with deeper levels nullable when not used. When fields is omitted, the endpoint returns every scalar column of the table; array (STRING_ARRAY) columns are excluded unless you request them explicitly in fields. For stable pagination, repeat the same request body and pass the previous response's nextPageToken as pageToken. Spend amount values and the currency column are in the tenant display currency (EUR when the tenant has none configured) unless the request sets currency; either way they are converted with the daily rate on or before transaction_date.

Starts an asynchronous export for a table. The returned export id can be used to poll status until the export is ready for download. Spend amount values and the currency column are in the tenant display currency (EUR when the tenant has none configured) unless the request sets currency; either way they are converted with the daily rate on or before transaction_date.

Polls the current status of an asynchronous export. When status is succeeded, the response includes time-limited download URLs. Three retention windows apply: the export status is available for 24 hours after the export was started, after which this endpoint returns 404; each download URL is valid for 1 hour and is re-issued on every poll; the exported files are deleted 7 days after the export. Start the export and download the files within the same 24-hour window.