List extraction results

Paginated extraction results for the project, with optional filtering, sorting, and search. Supplying exactly one agent_ids value also returns the document sources for each result.

On this page

HTTP request 

GET https://api.parsewise.ai/api/v1/projects/{project_id}/results/

URI Parameters

Name In Required Type Description
agent_ids query No string Comma-separated UUIDs. Restricts the response to rows produced by the listed agents (defaults to all of the project’s agents). Pass exactly one to also receive sources in the response.
dimension_order query No string Comma-separated dimension names controlling how rows are grouped and how columns are laid out across multiple agents. Order matters: earlier names group at the outermost level of the result table.
filter_inconsistent query No boolean When true, only return rows whose values are not fully consistent across the source documents. Defaults to false.
filter_missing query No boolean When true, only return rows whose resolved value is missing or empty. Defaults to false.
filter_review_status query No string Comma-separated review statuses to include. Accepted values: positive, negative, unreviewed. When omitted, no review-status filtering is applied.
filters query No string JSON-encoded dimension filter, applied to the rows. The value is a JSON object whose keys are dimension names (matching the name of a dimension on one of the project’s agents, e.g. "Document" or "Region") and whose values are arrays of dimension values to keep. Within a single key the values are OR-ed (a row matches the key if it has any of the listed values for that dimension); across keys they are AND-ed (a row must match every key to be returned). Empty value arrays and unknown dimension names are ignored. Example: {"Region": ["EMEA", "APAC"], "Year": ["2024"]}.
include_dimension_values query No boolean When false, omits the all_dimension_values field from the response to reduce payload size. Defaults to true.
page query No integer 1-indexed page number to return. Defaults to 1.
page_size query No integer Number of rows per page. Defaults to 100.
project_id path Yes string (uuid)  
search query No string Case-insensitive substring match against resolved values.
search_column query No string Agent name to restrict search to a single column. Ignored when search is not supplied.
sort_by query No string Column to sort by. Only applied when a single agent is selected and no dimension_order is supplied; otherwise the default ordering is used.
sort_order query No string Sort direction for sort_by. Defaults to asc.

Request Header

Name Required Type Description
X-API-Key Yes string API key with the pw_live_ prefix. See Authentication.

Responses

Status Type Description
200 V1ResultsListResponse —

Security

  • ApiKeyAuth — apiKey — in X-API-Key header. API key with pw_live_ prefix.

Python example 

import os
import requests

API_KEY = os.environ["PARSEWISE_API_KEY"]
BASE_URL = "https://api.parsewise.ai/api/v1"

project_id = "<uuid>"

params = {
    # "agent_ids": "...",  # Comma-separated UUIDs
    # "dimension_order": "...",  # Comma-separated dimension names controlling how rows are grouped and how columns are laid out across multiple agents
    # "filter_inconsistent": False,  # When `true`, only return rows whose values are not fully consistent across the source documents
    # "filter_missing": False,  # When `true`, only return rows whose resolved value is missing or empty
    # "filter_review_status": "...",  # Comma-separated review statuses to include
    # "filters": "...",  # JSON-encoded dimension filter, applied to the rows
    # "include_dimension_values": False,  # When `false`, omits the `all_dimension_values` field from the response to reduce payload size
    # "page": 0,  # 1-indexed page number to return
    # "page_size": 0,  # Number of rows per page
    # "search": "...",  # Case-insensitive substring match against resolved values
    # "search_column": "...",  # Agent name to restrict `search` to a single column
    # "sort_by": "...",  # Column to sort by
    # "sort_order": "asc",  # Sort direction for `sort_by`
}

resp = requests.get(
    f"{BASE_URL}/projects/{project_id}/results/",
    headers={"X-API-Key": API_KEY},
    params=params,
)
resp.raise_for_status()
print(resp.json() if resp.content else None)

Definitions

ExtractionStatusEnum

Type: string. One of:

  • Pending
  • Processed
  • No Result

LatestReviewStatusEnum

Type: string. One of:

  • none
  • positive
  • negative

NullEnum

Type: any. One of:

  • None

ResolutionStatusEnum

Type: string. One of:

  • Not resolved
  • Requires attention
  • Resolved
  • Ignored
  • No result

SourceType629Enum

Type: string. One of:

  • document
  • web
  • document_and_web

V1ResultDimensionInstance

One dimension instance (dimension + value) attached to an extraction result.

Name Required Type Description
id Yes string (uuid) Identifier of this dimension instance.
dimension_id Yes string (uuid) Identifier of the dimension this instance belongs to.
dimension_value Yes string Value of the dimension (e.g. “EMEA”, “2024”).
name Yes string Name of the dimension (e.g. “Region”).

V1ResultSource

Name Required Type Description
document_name Yes string Stored filename of the source document.
page_number Yes integer 1-indexed page number within the source document.
document_id Yes string (uuid) Stable identifier of the source document.
atomic_value_ids Yes array<string (uuid)> Identifiers of the atomic values whose resolutions cite this source.

V1ResultWebSupplement

Name Required Type Description
text Yes string Web-sourced supplementary text.
sources Yes array<V1ResultWebSupplementSource> Web pages cited for the supplement.

V1ResultWebSupplementSource

Name Required Type Description
url Yes string (uri) URL of the web source.
title Yes string Title of the web source.

V1ResultsListResponse

Name Required Type Description
results Yes array<V1ResultsListRow> Rows of the extraction result table on this page.
count Yes integer Number of rows returned on this page (i.e. the length of results). For the cross-page total, use total_rows.
page Yes integer 1-indexed page number of the returned page.
page_size Yes integer Requested page size (rows per page).
total_pages Yes integer Total number of pages available for the current filters.
total_rows Yes integer Total matching rows across all pages for the current filters.
displayed_dimensions Yes array<string> Ordered dimension names used to group/lay out rows in the response, reflecting the effective dimension_order.
agent_ids Yes array<string (uuid)> Agents included in the response.
recommended_dimension_order No array<string> Suggested dimension ordering for multi-agent views, or null when no specific ordering is recommended.
all_dimension_values No object<string, array<string>> Map of dimension name to the full set of values seen for that dimension across the matching rows, useful for building filter pickers. Omitted when include_dimension_values=false.
sources No array<V1ResultSource> Document pages cited by the results on this page. Only returned when exactly one agent_ids value is supplied.

V1ResultsListRow

Name Required Type Description
id Yes string (uuid) Identifier of the atomic value this row represents. To fetch the full detail for the resolved value, use the id from resolution_result with GET /results/{id}/.
agent_name Yes string Name of the agent that produced this row.
value_type Yes V1ResultsListRowValueTypeEnum Value type declared on the agent.
extraction_status Yes ExtractionStatusEnum Per-row extraction status.
resolution_result Yes V1RowResolutionResult Resolved value and metadata for this row (same shape as the GET /results/{id}/ response minus atomic_value_id, dimension_instances, and sources), or null if no resolution exists yet.
dimension_instances Yes array<V1ResultDimensionInstance> Dimension instances this row is keyed by (one entry per dimension configured on the agent).

V1ResultsListRowValueTypeEnum

Type: string. One of:

  • string
  • number
  • bool
  • date
  • datetime

V1RowResolutionResult

Resolved value and metadata for a row in the results list.

Same shape as the GET /results/{id}/ response minus atomic_value_id, dimension_instances, and sources.

Name Required Type Description
id Yes string (uuid) Stable identifier of the resolution result.
result_name Yes string Human-readable label for this result, typically a concatenation of its dimension values.
agent_name Yes string Name of the agent that produced this result.
value Yes any Resolved value. Shape depends on the agent’s value_type: string -> string; number -> number; bool -> boolean; date -> ISO 8601 date string YYYY-MM-DD; datetime -> ISO 8601 date-time string. May be null for missing values.
resolution_status Yes ResolutionStatusEnum Resolution outcome for this entity.
justification_for_resolution_status Yes string (nullable) AI-generated explanation of why this resolution status was chosen. Null when not applicable.
references Yes object<string, array<string | integer (nullable)>> Citation map keyed by reference_id (the same IDs referenced from justification_for_resolution_status and from extraction passages). Each value is a 2-element array [document_name, document_page]: document_page is the 1-based page number, or null for sources without pages (e.g. websites). Empty object {} when there are no citations.
source_type Yes SourceType629Enum Where the resolved value came from. document when the value was derived from uploaded documents only; web when it was derived from web-search results only; document_and_web when both contributed. The non-document values only appear when enable_web_search is set on the agent.
is_web_supplement Yes boolean True when this result was supplemented with web search data rather than derived purely from uploaded documents.
web_supplement No V1ResultWebSupplement Populated only when web search is enabled on the agent and web data contributed to the resolution; otherwise null.
updated_by_email Yes string (email) (nullable) Email of the last user to manually override this resolution, or null if never manually updated.
created_at Yes string (date-time) When the resolution was first produced.
updated_at Yes string (date-time) When the resolution was last updated.
latest_review_status No LatestReviewStatusEnum | NullEnum Review verdict from the latest review. Null when no review has ever been recorded.
latest_review_comment No string Free-text reviewer comment from the latest review. Empty string when no comment was given.
latest_review_at No string (date-time) (nullable) When the latest review was recorded. Null when no review has ever been recorded.
latest_review_by No string (email) (nullable) Email of the user who recorded the latest review. Null when unknown or no review exists.