Enterprises are increasingly integrating generative AI workloads, AI inferencing, and retrieval-augmented generation (RAG) workloads into their core business operations. And that shift was on full display at the recent Tech World @CES in Las Vegas, as Lenovo announced a suite of purpose-built enterprise servers, solutions, and services for AI inferencing workloads. The move marks a shift from training LLMs to leveraging them for stronger data insights.
Additionally, Kubernetes has become the backbone of modern infrastructure. According to a poll by Veeam, 52% of organizations are running containers in production with Kubernetes emerging as the go-to container orchestration platform for AI deployments.
As AI workloads become mission-critical, ensuring their resilience and protection is critical. Without purpose-built protection, recovery is complicated and costly. This is where the Lenovo AI Factory (Hybrid AI 285 platform) with resilience powered by Veeam Kasten comes into play.
What is the Lenovo AI Factory with Veeam Kasten?
The Lenovo AI Factory (Hybrid AI 285 platform) offers an enterprise-grade, scalable architecture for AI workloads featuring built-in data resilience using Veeam Kasten. The result delivers secure, Kubernetes-native data protection and application mobility at scale and across a wide range of distributions and platforms. Proven to recover entire applications simply, quickly, and reliably, Veeam Kasten gives operations and application teams the confidence needed to withstand the unexpected.
The joint solution embeds Kubernetes-native, application-centric protection directly into a GPU-optimized AI infrastructure stack. It enables rapid recovery, immutable backup, and automated, policy-driven protection for production AI workloads. Kasten’s core capabilities include backup and restore, disaster recovery, application mobility, and ransomware protection.
What Are the Benefits for Enterprises of Using Lenovo AI Factory with Veeam Kasten?
By combining Lenovo’s high-performance compute and enterprise storage platforms with Veeam Kasten’s Kubernetes-native protection capabilities, organizations can:
- Preserve immutable recovery points to strengthen ransomware resilience.
- Protect AI pipelines end-to-end across scalable units.
- Enable seamless application mobility across clusters and environments.
- Minimize downtime for inferencing and mission-critical AI services.
- Deploy AI infrastructure with resilience engineered from day one.
- Recover from AI prompt injection attacks and restore state for forensic analysis.
This approach allows organizations to confidently scale AI initiatives while maintaining operational continuity, cyber resilience, and performance integrity across mission-critical workloads.
Unlocking Opportunities While Overcoming the Challenges of Adopting AI
As AI environments transition from pilot to production, organizations encounter new data resilience risks. It’s estimated some 90% of enterprise initiatives fail because the data powering AI cannot be trusted. AI workloads are stateful and distributed, spanning various models, datasets, configurations, and types of metadata across Kubernetes clusters. Traditional backup solutions for virtual machines (VMs) lack the application awareness needed for cloud-native AI pipelines, and inferencing services demand low downtime and rapid recovery to maintain business continuity.
By embedding resilience directly into the Lenovo AI Factory platform, organizations can shift from reactive recovery to policy-driven AI protection, enable rapid model rollback and version control, and protect Kubernetes applications holistically. This validated design transforms AI resilience from an afterthought into a competitive advantage, ensuring that AI systems remain available, recoverable, and secure at enterprise scale.
Solution Components
The Lenovo Hybrid AI 285 platform scales from a single server with just 4 GPUs to a rack Scalable Unit (SU) with four servers and 32 GPUs, up to 5 SUs with 20 Servers and 160 GPUs. The main hardware components include storage from the Lenovo (DM7200F), compute with Lenovo ThinkSystem SR675 V3 GPU-rich servers, RTX Pro 6000 Blackwell Server Edition GPUs, and NVIDIA H200 NVL GPUs.
The software stack features Veeam Kasten, which provides Kubernetes-native data protection, backup, and restoration. It ensures safety and security for enterprise data, with encryption, granular access management, and Virtual Machine protection. Because Veeam Kasten is cloud-native, it is deployed on the Kubernetes cluster next to the AI, application, and VM resources it protects. This negates the need for dedicated compute or hardware for data protection while allowing data protection to be automated and integrated into CI/CD pipelines.
Veeam’s Commitment to Accelerating Safe AI at Scale
Recently, Veeam completed the acquisition of Securiti AI. The unified Veeam and Securiti AI platform gives customers real-time visibility, security, governance, and recovery across all data — production, backups, AI pipelines, cloud, and on–premises — with the speed of AI.
With this unified platform, organizations can:
- Understand data everywhere: Real-time Veeam Data Command Graph with classification, lineage, and continuous risk scoring.
- Secure data and AI: Leading Data Security Posture Management (DSPM), privacy, governance, and access controls combined with Veeam’s threat detection and identity-aware protection.
- Recover and rollback with precision: Cleanroom-validated restore up to five times faster, with granular rollback for datasets, embeddings, and model weights.
- Unleash trusted data for AI: Build governed pipelines and enterprise search in minutes, with entitlements, privacy, and policy enforced automatically.
Looking Ahead
The Lenovo Validated Design (LVD) for AI data resilience, powered by Veeam Kasten, offers a robust and scalable solution for protecting AI workloads. By integrating high-performance compute and storage platforms with Kubernetes-native protection, organizations can ensure their AI systems remain resilient, secure, and operationally efficient. This innovative approach addresses the challenges of AI resilience, providing a competitive edge in the ever-evolving AI landscape.
About Lenovo and Veeam
Through a strong integrated partnership, Lenovo and Veeam deliver a unified, modern backup and recovery solution that combines Veeam’s secure software with high-performance Lenovo hardware that empowers customers to minimize downtime, recover any workload anywhere, and helps ensure business continuity with confidence and peace of mind. Our joint solution also integrates Lenovo ThinkSystem data storage, Veeam Data Platform with Veeam Vault on a foundation of Lenovo infrastructure to deliver a multi-tiered data protection strategy.