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

# Document Generation & Parsing

> AI creates and parses leases, notices, and reports using your templates and data.

## What it does

* Auto-fills leases, notices, and summaries from templates and property data.
* Parses uploaded PDFs to extract structured fields for tenants, units, and payments.
* Maintains version history and auditability.

## Why it matters

* Save time on repetitive document prep.
* Reduce errors and rework with consistent templates.
* Keep structured data in sync with your source of truth.

## How it works

<Steps>
  <Step title="Ingest templates and docs">
    Upload templates (DOCX/PDF) and specify merge fields. Upload PDFs to parse (leases, receipts, inspections).
  </Step>

  <Step title="Generate outputs">
    AI fills templates with unit, tenant, and payment data; adds clauses based on rules; and produces PDFs.
  </Step>

  <Step title="Route and store">
    Send documents to tenants, vendors, or storage (Drive/S3). Store parsed fields for search and reporting.
  </Step>
</Steps>

## Configuration checklist

* Template fields mapped to properties, units, tenants, rent schedules, and clauses.
* Approval routing for generated documents.
* Storage destination and retention rules.

## Example generation request

<RequestExample>
  ```json Lease generation theme={null}
  {
    "template_id": "tpl_lease_v3",
    "tenant_id": "tenant_45",
    "unit_id": "unit_9A",
    "clauses": ["pet_addendum", "parking"],
    "output_format": "pdf"
  }
  ```
</RequestExample>

<ResponseExample>
  ```json Lease output theme={null}
  {
    "document_id": "doc_1209",
    "status": "generated",
    "url": "https://files.casaai.dev/leases/doc_1209.pdf",
    "checksum": "5f8c1d..."
  }
  ```
</ResponseExample>

## Diagram

```mermaid theme={null}
flowchart TD
  templates[Templates + PDFs] --> generate[AI fill + parse]
  generate --> outputs[Signed PDFs + structured fields]
  outputs --> storage[Storage (Drive/S3)]
  outputs --> webhooks[Webhooks / API consumers]
```