NAnace
Workspaces

Locate information

Names the pages and tables most likely to hold a piece of information. A decision model walks the workspace's files, sections and pages, built at ingestion, and answers with each piece, the read-file call that opens it and its text (the best pages in full unless pointer asks for opening lines only), then the tables to query. text is that answer as NDI's own search agent reads it; pages, tables and population hold the same as fields. Locate is stateless: send back the pieces earlier calls offered (exclude) and what you have read (already_read) to be shown something new. The first call after the workspace's files change reads them first and can take a minute or more; a call that times out keeps running, so send it again with the same Idempotency-Key. Limited to 150 per account every 10 minutes. Free.

post/v1/workspaces/{id}/tools/locate

Request

Path parameters

idstring required

The workspace's workspace_id.

Headers

Idempotency-Keystring

Up to 200 characters, unique per account. A call that timed out keeps running: send it again with the same key and body to get its answer instead of running it twice.

Request body

needstring required

What is still needed, in a few words: a year, an entity, a line item, a definition.

questionstring

The whole question the need is part of; every decision weighs the need within it. Leave it out and the need is the whole question.

kinteger

How many pieces to name.

everyboolean

The need names a kind of file and every one is wanted (each invoice, every statement): the model rates each file of that kind and lists the members in population.

path_prefixstring

Only files under this folder; paths adds files. Leave both out for the whole workspace.

pathsstring[]

Only these files (with path_prefix, its folder's files too). An unknown path, or one the key can't read, is refused.

pointerboolean

Name every page in text by its opening lines only. Leave it out and the best pages come in full, as a task's first locate wants them; a task's later locates send it, so page text does not pile up in an agent's context.

Response

The answer.

textstring required

The answer as NDI's own search agent reads it: each piece with the read-file call that opens it and its text (the best pages in full, the rest by their opening lines; every page by its opening lines with pointer), then a population's members and the tables to query.

population_unratedinteger

Files in folders rated likely to hold members that the budget ran out before rating: above 0, population is incomplete; narrow the need to a folder, kind or period and locate again with every.

job_idstring nullable

The job the call ran as; GET /v1/documents/jobs/{id} reads it.

Changes

No recorded changes to this operation across all 1 revision of this API.