As enterprises push artificial intelligence projects beyond pilots and into production, infrastructure limitations are becoming harder to ignore—and creating a new opening for channel partners.
Dennis Frank, Vice President, EMEA Strategic Partners & Alliances at Hitachi Vantara, spoke with Channel Insider about why storage, data pipelines, and governance are emerging as critical AI bottlenecks, how partners’ business models may need to change, and where solution providers can position themselves as customers build the infrastructure required to run AI at scale.
Enterprise AI exposes infrastructure and data gaps
What changes when enterprises move AI from pilots into production?
Taking AI from isolated experiments to enterprise-wide environments reveals critical gaps like:
- The inability of legacy storage architecture to provide AI-level throughput
- The skyrocketing cloud costs enterprises incur when their data volumes explode
- The struggle of on-premises arrays to perform as expected during training
Such barriers may require an enterprise to revisit its entire IT infrastructure even after it believed it was on the right track and ready to scale its AI. For channel partners, that creates both risk and opportunity: customers are looking for guidance, but they also need partners that can move beyond pilots and help build the data foundation required for AI.
Where are enterprises running into the biggest infrastructure challenges?
Many channel partners to date have focused primarily on procuring and deploying the best available GPU technology to support enterprise AI models. That’s important, but it fails to appreciate the underlying data problem many enterprises still face.
The true constraint is no longer the model or hardware; it’s the data pipeline, storage and network architecture, and the operational discipline required to run these systems reliably.
Most enterprise data is unstructured and unclassified, making it difficult to access, govern and use effectively for AI.
If you feel like you’ve heard this before, it’s because you have. This is essentially the same problem that channel partners and their customers ran into with cloud migration, ERP modernization and machine learning. But now AI is bringing it new urgency.
The bottom line? AI success hinges on high-quality data and AI-ready data infrastructure.
Data readiness becomes the AI scaling test
What does the data tell us about how prepared organizations really are for AI at scale?
According to a Gartner survey, 63% of organizations either lack or are unsure whether they have the right data management practices for AI, a gap that could put AI initiatives at risk as traditional data management approaches struggle to keep pace with AI’s demands.
Meanwhile, our own research reveals that:
- While 59% of organizations cite high-quality data as the top driver of AI success,
- Just 42% consider themselves data-mature and ready to support AI at scale, and
- Legacy data infrastructure is contributing to $108 billion in wasted global AI investment
Together, these findings show a practical reality for the channel: AI adoption is no longer just a software, model, or GPU conversation. It is a data infrastructure conversation.
AI pushes channel partners beyond transactional models
Do traditional channel business models still work for AI deployments?
Channel partners, like all of us, have faced an incredible amount of change in recent years. The rise of cloud platforms and as-a-service models led to countless articles and trade show sessions encouraging the channel to shift from transactional to consumption-based models.
Despite all that talk and IT industry evolution – and, now, the accelerating adoption of AI – channel partner incentives often still focus on CapEx cycles and point-in-time transactions.
But enterprises need flexible models that allow them to iterate as they expand AI adoption. AI initiatives rarely scale in a straight line.
Customers may start with a specific use case, expand to new data sets, adjust performance requirements, and rethink deployment models as business needs evolve. Channel partners need commercial models that can support that reality.
Governance, sovereignty reshape the partner opportunity
How are data governance and sovereignty requirements adding to the challenge?
With data sovereignty regulations and data governance expectations spreading like wildfire, the data management problem is becoming even more pronounced.
With sovereign data rules popping up almost everywhere in the world, the need for modern data infrastructure that allows enterprises to manage, control, and get visibility into where their data is stored is becoming even more critical.
What will separate the channel partners that succeed in the AI era from those that don’t?
Partners that cling to the past put both themselves and their customers at risk of marginalization and losing competitive ground.
The partners delivering these mission-critical capabilities will lead the market into the future. Those that can help customers modernize storage, improve data visibility, manage hybrid environments, and align infrastructure with AI outcomes will be best positioned as enterprises move from AI ambition to AI execution.
The AI infrastructure reckoning is here. For the channel, the question is whether partners will wait for customers to hit the wall, or help them build the foundation to scale AI with confidence.