# Apryse Server SDK

Process documents programmatically on-premises, or in any public or private cloud. The Apryse Server SDK supports the full document lifecycle for 30+ file types including PDF, Office, and CAD with conversion, editing, redaction, data extraction, digital signatures and more.

[View Server Docs](https://docs.apryse.com/core/guides/get-started)

## Secure Document Automation for the Modern Enterprise

Embed Apryse Server SDK to deliver scalable performance into your backend document processing and workflow applications

### Dynamic Document Generation

### File Conversion

### Smart Data Extraction

## Dynamic Document Generation

Accelerate document tasks by programmatically merging dynamic data with templates to produce thousands of personalized, brand-consistent documents like invoices and reports in seconds.

[View the Docs](https://docs.apryse.com/core/guides)

### File Conversion

Normalize diverse file types, including Office, CAD, email, and images into searchable PDFs without requiring native software, ensuring long-term compatibility and archiving.

[Read the Document Conversion Guide](https://apryse.com/document-conversion-guide)

### Smart Data Extraction

Transform unstructured documents into machine-readable insights by utilizing AI to extract structured data, including tables and key-value pairs, making the data accessible to power autonomous AI agents or simply store for analysis.

[Read the Smart Data Extraction Guide](https://apryse.com/smart-data-extraction-sdk-evaluation)

## **Server SDK FAQs**

##### **Is my data sent to an external server when using the Server SDK?**

No. One of the primary benefits of the Server SDK is that it is self-hosted. All document processing happens locally on your infrastructure. Your files never leave your network, making it an ideal choice for industries with strict privacy requirements like Finance, Healthcare, and Government.

##### **Which languages and frameworks are supported?**

We support nearly every major [backend environment](https://docs.apryse.com/core/guides/get-started). You can use the same core high-performance engine across:

- Frameworks: .NET 8/9, .NET Core, .NET Framework, and Node.js.
- Languages: C++, Java, Python, Go, PHP, Ruby, and Objective-C.
- Cloud & Mobile: Comprehensive support for AWS Lambda, Azure Functions.

##### **How does the Server SDK support AI agent development and deployment?**

The Server SDK serves as the [critical data preparation layer](https://apryse.com/apryse.com/AI-at-Apryse) for AI ecosystems, specifically supporting the development of Retrieval-Augmented Generation (RAG) pipelines and autonomous AI agents. By converting unstructured PDFs into clean, labeled JSON, the SDK provides the high-quality, structured "ground truth" that Large Language Models (LLMs) require to function without hallucinating. Because the SDK is designed for deployment on-premises or in private clouds, AI agents can be deployed in air-gapped or highly regulated environments where sending sensitive document data to external third-party extraction APIs is not an option.

##### **Does the Server SDK require Microsoft Office to be installed to convert documents?**

No. Apryse uses a proprietary rendering engine [that Office](https://apryse.com/capabilities/conversion/pdf-to-office) documents (Word, Excel, PowerPoint) to PDF directly. This allows you to run high-fidelity conversions on Linux servers or Docker containers without the overhead or licensing costs of MS Office.

##### **How is Smart Data Extraction different from OCR?**

While traditional Optical Character Recognition (OCR) is effective for simple text retrieval, the Apryse Server SDK goes significantly further by utilizing [Smart Data Extraction](https://apryse.com/capabilities/smart-data-extraction) to understand the underlying structure of PDFs. This AI-enabled technology moves beyond simple character identification to reconstruct the semantic logic of a file, effectively turning documents into structured, actionable data. Smart Data Extraction can autonomously detect complex table structures, including nested cells and headers, and accurately map key-value pairs, such as linking an invoice number directly to its corresponding value.