Reinsurance Submission Overview

Read reinsurance workbooks cell by cell, map labels into a client target structure, check the figures, and export the facts with the /v1/submissions API.

The Reinsurance Submission app reads every reinsurance workbook of a workspace cell by cell, maps the labels it finds into a client target structure, checks the figures against each other, and exports the facts in the client's own layout. A cedent sends a workbook of loss triangles, premium tables and exposure schedules; the app turns each one into rows of facts, each carrying a metric, a year, the field values that qualify it, and the exact sheet and cell it was read from. The /v1/submissions routes expose the same service the app screen uses.

These endpoints belong to the Reinsurance Submission app (catalog slug submissions). A workspace uses them after adding the app; GET /v1/apps lists the workspace's apps by slug. A submission is a top-level workbook document (.xlsx, .xlsm or .xls) of the workspace. Workbooks embedded inside another workbook belong to their parent's capture and are not listed on their own.

The working loop has five steps. Capture reads a workbook into tables of cell values. Map projects those tables into facts under one target structure (a spec_id). Review settles the label mappings: rule-made mappings are confirmed without a model call, and an agentic review confirms, fixes or asks about the rest, leaving only a few questions for the client's experts. Check compares figures with the totals and identities they should add up to. Export returns the facts in the structure's own columns, order and names.

Authentication uses a tlnc_ API key or an OAuth token in the Authorization: Bearer header. Reads need the read scope. Every write (POST /v1/submissions/process, /review, /questions/answer and /:documentId/fields) needs the write scope and the key's person to be a member or above in the workspace; a key created by a Viewer is refused with 403 insufficient_role. Every response of this surface wraps its payload in a data field.

Worked Example: From Workbooks to Exported Facts

List the workspace's target structures and pick one, then capture and map every workbook under it. A dry run first returns the plan without running anything.

curl "https://api.talonic.com/v1/submissions/specs" \
  -H "Authorization: Bearer $TALONIC_API_KEY"

curl -X POST "https://api.talonic.com/v1/submissions/process" \
  -H "Authorization: Bearer $TALONIC_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"spec_id": "5d1c0a52-0c1e-4b64-9d2e-7a3c4f1e9b10", "dry_run": true}'

curl -X POST "https://api.talonic.com/v1/submissions/process" \
  -H "Authorization: Bearer $TALONIC_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"spec_id": "5d1c0a52-0c1e-4b64-9d2e-7a3c4f1e9b10"}'

Then start the agentic review of the structure's label mappings, poll its status, and answer whatever questions remain. When the open questions are settled, export the facts.

curl -X POST "https://api.talonic.com/v1/submissions/review" \
  -H "Authorization: Bearer $TALONIC_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"spec_id": "5d1c0a52-0c1e-4b64-9d2e-7a3c4f1e9b10"}'

curl "https://api.talonic.com/v1/submissions/review?spec_id=5d1c0a52-0c1e-4b64-9d2e-7a3c4f1e9b10" \
  -H "Authorization: Bearer $TALONIC_API_KEY"

curl "https://api.talonic.com/v1/submissions/export?spec_id=5d1c0a52-0c1e-4b64-9d2e-7a3c4f1e9b10" \
  -H "Authorization: Bearer $TALONIC_API_KEY"

{
  "data": {
    "name": "Client long format",
    "header": ["Metric", "Year", "LoB Client", "Figure", "Document", "Source"],
    "rows": [
      ["Incurred Loss", "2023", "Homeowners", 1250000, "utica-first-2024.xlsx", "Triangles!D14"]
    ]
  }
}