Read files and ask questions - NDI Platform API

These tools are inline (not jobs) except where noted. They read ingested workspace files.

Folder and file metadata

Python
TypeScript
curl

from ndi_sdk import NdiClient

with NdiClient() as client:
    workspace_id = "550e8400-e29b-41d4-a716-446655440001"
    folder = client.tools.folder_metadata(workspace_id, directory="reports/")
    print(folder)
    meta = client.tools.file_metadata(workspace_id, path="reports/report.pdf")
    print(meta)

import { NdiClient } from "ndi-sdk";

const client = new NdiClient();
const workspaceId = "550e8400-e29b-41d4-a716-446655440001";
const folder = await client.tools.folderMetadata(workspaceId, { directory: "reports/" });
const meta = await client.tools.fileMetadata(workspaceId, { path: "reports/report.pdf" });
console.log(folder, meta);

curl -s -X POST "$NDI_BASE_URL/v1/workspaces/$WS/tools/folder-metadata" \
  -H "X-API-Key: $NDI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"directory":"reports/"}'

curl -s -X POST "$NDI_BASE_URL/v1/workspaces/$WS/tools/file-metadata" \
  -H "X-API-Key: $NDI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"path":"reports/report.pdf"}'

folder_metadata is an offset page (start_after / page_size), not a cursor list. See Pagination.

Read file

Returns extracted text for a path. Optional pages and max_chars.

from ndi_sdk import NdiClient

with NdiClient() as client:
    workspace_id = "550e8400-e29b-41d4-a716-446655440001"
    page = client.tools.read_file(
        workspace_id,
        path="reports/report.pdf",
        pages=[1],
        max_chars=4000,
    )
    print(page.content)

QA file

Ask a natural-language question of one uploaded document, image, or recording. Do not use it for Excel — use query tables.

from ndi_sdk import NdiClient

with NdiClient() as client:
    workspace_id = "550e8400-e29b-41d4-a716-446655440001"
    answer = client.tools.qa_file(
        workspace_id,
        path="reports/report.pdf",
        query="What were total liabilities at year end?",
    )
    print(answer.answer)
    for citation in answer.citations or []:
        print(citation)

Errors

file_not_ingested — ingest first
invalid_request — qa_file on a spreadsheet path