Table of Contents

Introduction

Why the data foundation matters

Expanding beyond collaboration-centric data

The zero‑copy data foundation pattern

Why OneLake changes the model

What this enables

Key takeaway

Learn more

Introduction

Enterprise AI is advancing rapidly, but many organizations are discovering a common challenge: their most valuable data is not accessible to the systems they want to use.

While collaboration platforms and modern data services are designed for AI integration, the majority of enterprise knowledge still resides in file systems, powering critical applications, operational workflows, and long-standing systems of record. These environments were built for performance, reliability, and scale, not for AI-driven discovery or reasoning.

This creates a fundamental gap. Organizations have the data they need to power AI, but that data often remains outside the reach of modern AI platforms.

In this blog series, we explore how to close that gap without rewriting workflows, migrating data, or disrupting existing systems.

Co-authors: 

• Thomas Willingham, Azure NetApp Files Product Manager 
• Sean Luce, Azure NetApp Files Product Manager 

Why the data foundation matters

As organizations accelerate their adoption of AI, a familiar challenge keeps emerging: AI is only as effective as the data it can access. While models and tools continue to advance rapidly, many enterprises are constrained by data foundations that were never designed for AI workloads.

The issue is not a lack of data. In fact, most organizations already store their most valuable institutional knowledge on enterprise file systems:

• Engineering specifications and design documents
• Legal contracts and compliance records
• Policies and procedures
• Technical runbooks
• Historical reports and archives

For many enterprises, this content lives on high‑performance file storage such as Azure NetApp Files, accessed through long‑standing SMB and NFS workflows. The data is fast, reliable, and trusted, yet largely invisible to AI.

Files are rich in information, but they are fundamentally passive. Users must already know where to look, and AI systems cannot reason over data they cannot access. This disconnect between where enterprise content lives and where AI knowledge can operate is now one of the biggest blockers to scalable AI adoption.

Expanding beyond collaboration-centric data

Many AI scenarios today focus on collaboration-centric data that is:

• Cloud‑native
• Stored in SharePoint or OneDrive
• Well suited for small to medium‑sized repositories

This is an important and valuable class of data. But it represents only part of the enterprise reality. Enterprise AI requires extending beyond collaboration data to include systems of record, where the majority of institutional knowledge resides.

Most large organizations also rely on:

• Large NAS estates measured in tens or hundreds of terabytes, and often millions of files
• Hybrid and on‑premises environments
• Long‑standing SMB/NFS workflows that cannot be easily re‑platformed or migrated

These environments are not failures of SharePoint, they serve different needs. File systems remain the system of record for many regulated, performance‑sensitive, and document‑heavy workloads.

The challenge has been extending AI, and data platforms, to this class of enterprise file data without forcing migration, duplication, or workflow change.

This is the gap that Azure NetApp Files and Microsoft OneLake fill together.

The zero‑copy data foundation pattern

To unlock AI across enterprise file data, we need a different foundation, one that respects existing systems of record while making data accessible to modern AI services.

This solution introduces a zero‑copy, zero‑migration data foundation built on three key components:

Files continue to be created, updated, and managed on Azure NetApp Files exactly as they are today.

The Azure NetApp Files object REST API provides object‑based access to that data, enabling modern analytics and AI platforms to work directly with existing file datasets without copying, moving, or restructuring them.

This same zero-copy pattern is already used to unlock advanced analytics and AI scenarios with services such as Azure Databricks and Microsoft Fabric, validating its suitability for large‑scale enterprise AI workloads.

Why OneLake changes the model

Microsoft OneLake is often described as a “single data lake for the entire organization,” but its most important capability in this architecture is what it does not do.

OneLake does not ingest Azure NetApp Files data.

Instead, it:

• creates shortcuts as virtual pointers to external data,
• preserves ownership, permissions, and access patterns,
• avoids storage duplication and ETL pipelines,
• keeps Azure NetApp Files as the authoritative system of record.

This shortcut‑based model allows OneLake to unify data across clouds, platforms, and on‑premises environments while maintaining a single logical namespace.

From an AI perspective, this is critical. It means enterprise file data can be discovered, indexed, and reasoned over by downstream AI services without breaking compliance boundaries, duplicating storage, or introducing synchronization risk.

What this enables

With Azure NetApp Files connected to OneLake through the object REST API:

• enterprise file data becomes AI‑addressable,
• existing SMB/NFS workflows remain unchanged,
• data stays governed, secured, and owned by the original platform,
• AI systems gain visibility into previously inaccessible knowledge.

This architecture does not replace SharePoint‑based scenarios. Instead, it expands AI and data analytics scenarios across a broader, more representative enterprise data estate.

Key takeaway

Azure NetApp Files is the high-performance, intelligent, and cyber-resilient foundation for enterprise file data, powering AI with enterprise application file data wherever it lives.

By combining Azure NetApp Files, the object REST API, and Microsoft OneLake, organizations establish a unified, zero‑copy foundation that brings enterprise file data into scope for modern AI systems.

With the data foundation in place, the next step is turning that data into usable knowledge.

This AI‑ready data foundation enables higher‑level AI capabilities and agents that are governed and surfaced to users in later layers of the architecture.

In Part 2, From File Data to AI‑Powered Knowledge Pipelines using Azure NetApp Files object REST API we’ll explore how this foundation enables a knowledge pipeline using Azure AI Search and Azure OpenAI to support Retrieval-Augmented Generation (RAG).

Learn more

• From File Data to AI‑Powered Knowledge Pipelines using Azure NetApp Files object REST API | Microsoft Community Hub
• Bringing Enterprise File Data to Users with Azure NetApp Files, Microsoft Foundry, and M365 Copilot | Microsoft Community Hub
• Understand Azure NetApp Files object REST API | Microsoft Learn
• Configure object REST API in Azure NetApp Files | Microsoft Learn
• Connect OneLake to an Azure NetApp Files volume using object REST API | Microsoft Learn
• How Azure NetApp Files Object REST API powers Azure and ISV Data & AI services – on YOUR data | Microsoft Community Hub
• Unlocking Advanced Data Analytics & AI with Azure NetApp Files object REST API | Microsoft Community Hub