> ## Documentation Index
> Fetch the complete documentation index at: https://docs.aiinsurance.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Overview

> Extract structured data from unstructured insurance documents

<Note>
  **Planned** — Submission parsing endpoints are being added to the v1 API. See the [Roadmap](/api-v1-reference/roadmap) for overall API direction.
</Note>

The Submission Parsing API will provide AI-powered extraction of structured data from unstructured insurance documents. Upload an application, loss run, or other document and receive structured field data that can be fed directly into the transaction APIs.

## Key Concepts

* **Application parsing** — extract structured quote fields from ACORD forms, custom applications, and other submission documents
* **Loss history parsing** — extract structured loss history from loss run documents, focused on tabular financial data
* **Confidence scores** — each extracted field includes a confidence score so your workflow can decide what needs manual review

## Planned Endpoints

| Method | Endpoint                             | Description                                           |
| ------ | ------------------------------------ | ----------------------------------------------------- |
| POST   | `/v1/submissions/parse/application`  | Parse an application into structured quote fields     |
| POST   | `/v1/submissions/parse/loss-history` | Parse loss run documents into structured loss history |
| POST   | `/v1/submissions/parse`              | Best-effort parse of any document type                |

## How It Will Work

### Application Parsing

1. Upload an application document (ACORD form, custom application PDF)
2. The system extracts structured field data matching your company's configured quote fields
3. The response includes extracted fields with confidence scores
4. Feed the results into `POST /v1/quotes` to create a pre-populated quote, or into the policy transaction APIs for endorsements and renewals

### Loss History Parsing

1. Upload a loss run document
2. The system extracts structured loss history focused on tabular financial data
3. The response includes extracted loss history with confidence scores
4. Feed the results into a submission for underwriting review

### General Parsing

1. Upload any insurance document
2. The system determines the document type and extracts whatever structured data it can
3. Feed the results into a submission for review and processing

## Example: Automated Intake Workflow

Submission parsing composes with the rest of the v1 API to support automated intake:

1. **Parse** an incoming application → structured fields
2. **Create a quote** from the parsed fields → `POST /v1/quotes`
3. **Rate** the quote → `POST /v1/quotes/{quoteId}/rate`
4. **Review** — use confidence scores to flag fields for manual review
5. **Bind** → `POST /v1/quotes/{quoteId}/bind`

## Related Resources

* [Quotes API](/api-v1-reference/quotes/overview) — create quotes from parsed application data
* [Events API](/api-v1-reference/events/overview) — create events from parsed loss run data
* [Roadmap](/api-v1-reference/roadmap) — overall API direction