Partner POV | Unlocking AI potential: The essential role of hybrid multicloud strategies

This article was written by Jonsi Stefansson, CTO and Senior Vice President at NetApp, and was contributed by our partner, NetApp.

At NetApp, we recognize that AI is not merely a technological tool—it's a transformative mindset that can reshape organizations and industries. To harness the full potential of AI, your organization must cultivate a data-driven culture that permeates every level of your company.  

NetApp is not alone in adopting an AI mindset. Notably, hyperscalers are making substantial investments in AI and predictive analytics. Their role can be crucial in helping your business improve customer experiences and create new revenue streams through AI-driven innovations. 

However, each cloud provider offers distinct advantages for AI workloads, making a multicloud strategy vital:  

• AWS provides diverse pretrained models for various generative tasks, including image, text, and music creation.
• Google Cloud is making strides in developing specialized AI models, such as those tailored for healthcare applications like ultrasound image interpretation.
• Azure's generative AI (GenAI) solutions integrate seamlessly with Microsoft's ecosystem, offering a cohesive experience for organizations that are heavily invested in Microsoft products.

With NetApp® first-party, cloud-native storage solutions, your organization can quickly benefit from the hyperscalers' AI investments. For example, NetApp BlueXP™ workload factory for AWS integrates data from Amazon FSx for NetApp ONTAP with Amazon Bedrock foundational models, enabling the creation of customized retrieval-augmented generation (RAG) chatbots. This integration enables your organization to use your proprietary data in GenAI applications, enhancing the relevance and accuracy of AI-generated responses. 

Enhance flexibility, security, and compliance with multicloud

By using a multicloud approach, your business can take advantage of each cloud provider's unique strengths and choose the best platform for each GenAI RAG-based project, without being limited to just one provider's ecosystem. 

Moreover, multicloud data solutions are essential for complying with regulatory frameworks. For example, the Digital Operational Resilience Act (DORA) from the European Union goes into effect in January 2025. DORA security requirements apply to a wide range of financial institutions, including banks, investment firms, payment service providers, asset managers, and crypto-asset service providers. It also encompasses third-party information and communications technology (ICT) providers who deliver critical services to these financial organizations, such as data analytics platforms, software vendors, and cloud service providers. 

DORA requires financial firms to have strategies in place to manage the risk that's related to their third-party service providers, such as AWS and Microsoft Azure. This requirement applies whether it's a managed process such as an exit strategy or an unexpected event like a cyberattack. By using intelligent data infrastructure from NetApp, financial institutions can securely end contracts with third-party providers and seamlessly transfer training and inferencing data to a new cloud platform. This ensures uninterrupted business operations during the transition, maintains service quality for customers, and adheres to regulatory requirements. In addition, they can actively detect and safeguard their data, enabling rapid recovery if an attack occurs. 

Use hybrid cloud to protect sensitive data

Many businesses will choose public cloud services for AI. But NetApp believes that there are compelling reasons why specific organizations may decide to run AI workloads in their private data centers or to use a hybrid cloud model. For particular industries, such as healthcare, defense contracting, government, and finance, the sensitivity of their business data makes cloud-based data preparation, model training and fine-tuning, and inferencing unsuitable.  

NetApp data solutions support companies that opt for a do-it-yourself approach with proprietary or open-source models. They also support organizations that take advantage of a turnkey converged AI solution such as NetApp AIPod™ with Lenovo or FlexPod® for AI. And NetApp solutions are optimal for companies that adopt a hybrid model that combines data center resources with cloud-based services. 

NetApp data solutions support a hybrid multicloud strategy

AI has advanced rapidly, with models increasing in complexity, datasets expanding in size, and the demand for real-time insights becoming crucial. Your organization can use hybrid multicloud strategies to distribute your AI workloads across on-premises and different cloud environments, optimizing performance, cost, and resource allocation. And NetApp has the tools that you need to make your hybrid multicloud AI deployments a success. 

Unified data management. It's no secret that data silos slow down AI projects. NetApp intelligent data infrastructure unifies access to file, block, and object storage, offering configurations ranging from high-performance flash to cost-efficient hybrid flash storage. And it is available in data centers, in colocation facilities, and through our public cloud partners. NetApp is the only provider who offers first-party, cloud-native storage solutions on all three major public clouds: Amazon FSx for NetApp ONTAP, Azure NetApp Files, and Google Cloud NetApp Volumes. Your organization can easily move, manage, and protect your data across various cloud platforms, minimizing data silos and enabling your AI teams to access data when and where they need it. 

Integrated AI service capabilities. To make full use of the unique strengths of each cloud platform with industry-specific knowledge or business-specific information, your organization needs to integrate proprietary enterprise data with custom task-based models.  

NetApp has developed a variety of integrated toolkits that help you overcome this challenge: 

• With the launch of the GenAI capability in BlueXP workload factory, our AWS customers can deploy and manage RAG pipelines. This new capability enables customers to securely connect data in NetApp ONTAP® based storage with Amazon Bedrock to develop GenAI applications without having to copy it to Amazon S3.
• The NetApp GenAI Toolkit, which supports Google Cloud NetApp Volumes, speeds up the implementation of RAG operations while enabling secure and automated workflows that connect data stored in NetApp Volumes with Google Cloud Vertex AI.
• The NetApp GenAI Toolkit is also in preview in Azure NetApp Files in Microsoft Azure.

Data governance. Data classification involves categorizing data based on its sensitivity, value, and regulatory requirements across multiple clouds. Our comprehensive set of features goes beyond basic data cataloging. By using AI, machine learning (ML), and natural language processing technologies, BlueXP classification can categorize and classify data by type, redundancy, and sensitivity, highlighting potential compliance exposures. NetApp offers a range of data classification strategies that are tailored to the unique challenges that GenAI poses: 

• Data estate visibility. Improve the cleanliness of your data and gain knowledge about sensitive information with complete visibility of your entire NetApp data estate, both on premises and in the public cloud.
• Discovery of personal and sensitive data. Our classification capabilities can identify personally identifiable information (PII), credit card numbers, Social Security numbers, and bank account numbers. They also identify sensitive personal data such as health details, ethnic background, and sexual orientation. These capabilities facilitate compliance with regulatory requirements across jurisdictions.
• Data optimization. To reduce overhead and to confirm that AI models receive the most current context, you can eliminate duplicate, stale, and nonbusiness data that can distort results.

Security and compliance. NetApp's security features are highly effective in protecting your data and in maintaining compliance across multiple cloud environments. The built-in ransomware protection makes workload defense and recovery easier and faster. By using a single control plane, you can monitor your workload and assess data risk with customized recommendations from preconfigured protection policies. This capability helps in maintaining compliance with regulations such as the General Data Protection Regulation (GDPR) and the Health Insurance Portability and Accountability Act (HIPAA). In addition, your data can be automatically sent to a security information and event management (SIEM) system for efficient threat analysis and detection. 

Partnerships. In addition to our work with hyperscalers, we have forged strategic alliances with leading independent software vendors (ISVs) in predictive and GenAI, such as Dremio, PyTorch, Domino Data Lab, Run.ai, and Red Hat OpenShift AI. These partnerships support a hybrid multicloud strategy and enable NetApp to provide intelligent data infrastructure with state-of-the-art unified lakehouse platforms, AI cluster engines, and scalable AI and ML platforms.  

Take control of AI transformation

The adoption of hybrid multicloud strategies is essential for your organization to unlock the full potential of AI. An intelligent data infrastructure helps your teams manage data seamlessly across on-premises, hybrid, and multicloud environments. It facilitates access to the best resources for the AI project at hand while helping your organization maintain data security and compliance. And as AI continues to advance, your business must embrace a data-driven culture and use intelligent data infrastructure to stay competitive and to innovative in your industry.  

Industry-leading NetApp hybrid multicloud solutions can find the optimal location for your data at any time, empowering you to get the most out of your AI journey.