> For the complete documentation index, see [llms.txt](https://docs.uptiq.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.uptiq.ai/document-ai/document-ai-1/readme.md).

# Introduction

UPTIQ Document AI is the document-processing layer behind UPTIQ Solutions. It turns raw files into structured, reviewable, and actionable data. Teams use it to classify incoming documents, extract fields, detect fraud signals, generate outputs, measure accuracy, and send results to downstream systems.

Document AI supports document-heavy workflows across commercial lending, SBA, CRE, consumer lending, equipment finance, and digital banking. It combines OCR, layout analysis, LLM-based reasoning, configurable rules, and human review where precision matters.

### What Document AI does

Document AI covers the full document lifecycle:

* **Understand the file** — identify what the document is with [Document Classification](/document-ai/document-ai-1/document-classification.md).
* **Turn content into data** — extract fields, tables, and layout with [Document Extraction](/document-ai/document-ai-1/document-extraction.md).
* **Screen for risk** — apply configurable checks with [Fraud Detection](/document-ai/document-ai-1/fraud-detection.md).
* **Produce outputs** — review and download generated results with [Document Generation](/document-ai/document-ai-1/document-generation.md).
* **Validate quality** — benchmark extraction accuracy in [Test Dashboard](/document-ai/document-ai-1/test-dashboard.md).
* **Connect downstream systems** — push events and results through [Webhooks & Integrations](/document-ai/document-ai-1/webhooks-and-integrations.md).

### How the workflow fits together

Most teams use Document AI in this sequence:

1. **Ingest documents** in [Workspace](/document-ai/document-ai-1/workspace.md) or through an integrated pipeline.
2. **Classify** each file when the type is unknown or mixed.
3. **Extract** structured data from the confirmed document type.
4. **Validate** results with fraud rules, human review, or test coverage.
5. **Deliver** data and outputs to operations teams or connected systems.

This flow lets teams reduce manual review without losing control over accuracy, routing, or auditability.

### Core building blocks

Document AI is organized around a few core concepts:

* **Document types** define what the system can classify and extract. See [Document Types](/document-ai/document-ai-1/document-types.md).
* **Workspace** gives teams a shared operating view for activity, collections, and quick actions. See [Workspace](/document-ai/document-ai-1/workspace.md).
* **Users and roles** control who can configure, review, and administer the platform. See [User Management](/document-ai/document-ai-1/user-management.md).

### Who it's for

Document AI supports teams that need speed, accuracy, and control across document-heavy workflows:

* **Operations teams** — organize, process, and review documents at scale.
* **Risk and underwriting teams** — validate extracted data, detect issues, and support decisions.
* **Engineering teams** — connect outputs to internal systems, workflows, and event-driven processes.

### Typical use cases

Teams use Document AI for common workflows such as:

* **Intake and document triage** — identify incoming files and route them to the right workflow.
* **Financial spreading and underwriting prep** — extract financial data into structured, reviewable outputs.
* **Fraud screening and exception handling** — flag risky documents and focus reviewers on outliers.

### Workflow summary

Document AI follows a simple operating model:

`Ingest → Classify → Extract → Validate → Deliver`

* **Ingest** documents through Workspace or an integrated pipeline.
* **Classify** unknown or mixed files before processing.
* **Extract** structured fields, tables, and layout data.
* **Validate** with fraud rules, human review, and test coverage.
* **Deliver** outputs to downstream teams and systems.

### Where to start

Choose the page that matches your job to be done:

* Start in [Workspace](/document-ai/document-ai-1/workspace.md) to navigate day-to-day operations.
* Use [Document Classification](/document-ai/document-ai-1/document-classification.md) and [Document Extraction](/document-ai/document-ai-1/document-extraction.md) to process documents.
* Use [Fraud Detection](/document-ai/document-ai-1/fraud-detection.md), [Test Dashboard](/document-ai/document-ai-1/test-dashboard.md), and [Webhooks & Integrations](/document-ai/document-ai-1/webhooks-and-integrations.md) to validate and operationalize results.

### Frequently asked questions

<details>

<summary>What product capabilities does Document AI include?</summary>

Document AI supports classification, extraction, fraud screening, document generation, accuracy testing, and webhook-based delivery. It gives teams one workflow for turning raw documents into structured outputs and operational actions.

</details>

<details>

<summary>Can Document AI classify documents before extraction?</summary>

Yes. [Document Classification](/document-ai/document-ai-1/document-classification.md) identifies document type and category before extraction. Teams use it to route unknown files, mixed files, and unlabeled inboxes into the right extraction flow.

</details>

<details>

<summary>What kinds of documents can Document AI handle?</summary>

Document AI uses a shared registry of 115 document types across classification and extraction. Coverage includes financial statements, tax forms, bank statements, credit reports, identity documents, real estate documents, and more. See [Document Types](/document-ai/document-ai-1/document-types.md).

</details>

<details>

<summary>Can it process batches, ZIP files, or mixed PDFs?</summary>

Yes. Document AI supports single documents, batch uploads, ZIP archives, and multi-document PDFs. Classification can segment longer PDFs before extraction, and extraction can process multi-document files as part of the workflow.

</details>

<details>

<summary>Does Document AI only extract text?</summary>

No. [Document Extraction](/document-ai/document-ai-1/document-extraction.md) returns structured fields, tables, and layout data. That gives teams both machine-readable outputs and reviewable context against the source document.

</details>

<details>

<summary>Can teams review and correct extraction results?</summary>

Yes. Users can review extracted values against the source document and submit corrections. Some document types also support Chart of Accounts review for balance sheets, profit and loss statements, and property operating statements.

</details>

<details>

<summary>Does Document AI support fraud detection?</summary>

Yes. [Fraud Detection](/document-ai/document-ai-1/fraud-detection.md) runs pre-OCR visual checks and post-extraction rule-based checks. Teams can use recommended templates, natural-language rules, and custom logic to score document risk.

</details>

<details>

<summary>Can it generate finished reports or outputs?</summary>

Yes. [Document Generation](/document-ai/document-ai-1/document-generation.md) turns structured section data into downloadable DOCX reports. It is designed for outputs such as credit memos and financial summaries.

</details>

<details>

<summary>How do teams measure accuracy before going live?</summary>

Use [Test Dashboard](/document-ai/document-ai-1/test-dashboard.md). It benchmarks extraction results against ground truth, tracks field-level assertions, and rolls results into document-type and fleet-wide accuracy views.

</details>

<details>

<summary>Can Document AI integrate with downstream systems?</summary>

Yes. [Webhooks & Integrations](/document-ai/document-ai-1/webhooks-and-integrations.md) sends signed webhook events when classification, extraction, or generation jobs complete. Teams can monitor deliveries, retry failures, and retrigger events when needed.

</details>

### Related pages

* [Glossary](/document-ai/document-ai-1/glossary.md) for shared terms and workflow language.
* [Document Types](/document-ai/document-ai-1/document-types.md) for supported schemas and type coverage.
* [User Management](/document-ai/document-ai-1/user-management.md) for roles, access, and account controls.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://docs.uptiq.ai/document-ai/document-ai-1/readme.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.