Enterprises are racing to scale AI, yet ambition outpaces operational readiness. In March 2025, LTIMindtree commissioned Forrester Consulting to survey 576 IT and AI decision-makers across North America, Europe, and the Nordics.
Findings show critical computing, storage, networking, and data preparation gaps, while privacy, governance, and skills constrain deployment.
The study suggests adopting workload-aware architectures, hybrid-cloud strategies, strong data foundations, and trust-by-design governance. Combining these with partner co-engineering can help transform pilots into AI systems that are scalable, cost-efficient, and responsible, ensuring AI remains the primary driver of future-ready enterprise strategy.
Key Findings
Foundation Gaps Undermine Model Quality
Only 18% of the IT and AI decision-makers surveyed say their infrastructure adequately supports data preparation. This inadequacy degrades model quality and reliability by increasing data errors, extending preprocessing time, and reducing repeatable training results.
Limited Skills and Trust Slow Adoption
Just 30% of survey respondents report strong AI literacy or formal training programs inside their organizations. Meanwhile, 52% cite privacy and security concerns. Together, these gaps reduce deployment confidence and delay production rollouts.
Hybrid-First Data Architectures Are Gaining Ground
Most organizations still use traditional relational databases, data warehouses, and data lakes to store information. However, lakehouse patterns are emerging because they combine the governance of warehouses with the scale and flexibility of lakes, making them better suited for ML pipelines.
Data Throughput and Storage Performance Remain Bottlenecks
Solid-state drives (SSDs) are standard, reflecting routine storage upgrades. However, only 23% of respondents use high-speed parallel file systems, and 18% use Non-Volatile Memory Express (NVMe). These shortfalls limit data pipeline throughput, slow batch training, and real-time feature delivery.
Strategic Partners Are Critical to Move from Pilot to Production
Survey respondents identify technology and consulting partners as essential for co-engineering workload-aware architectures, implementing governance, and accelerating pilots into production. Partner engagement reduces execution risk, shortens time-to-value, and helps organizations align technical choices with business outcomes.
Unlocking scalable enterprise AI with robust infrastructure
moderated by Charlie Dai, Vice President and Principal Analyst at Forrester, featured industry leaders discussing the critical need for modern, agile, and robust IT infrastructure to support rapid AI workload growth. Krishnan Iyer (Chief Growth Officer, LTIMindtree), Pandiya Kumar Rajamony (Global Head of CIS Service Line), and Manoj GS (Head of CIS Practice, Americas) shared insights on managing trust, integration, and the evolving partner ecosystem. Klaus Glatz (CIO, Exyte) and Matthew Rogers (Field CTO, VAST Data) discussed practical architecture patterns and trade-offs, emphasizing the importance of specialized technologies and strategic approaches for successful enterprise AI adoption.
What You’ll Learn from This Study
- Why AI ambitions are outpacing operational readiness in global enterprises
- The top infrastructure gaps slowing down AI at scale: compute, storage, network, and data foundations
- How hybrid cloud and specialized architectures are becoming the new default for AI workloads
- The persistent trust barrier and what leaders are doing to overcome it
- Emerging investment priorities in AI platforms, cloud-native solutions, and data modernization
- Forrester’s recommendations to future-proof your infrastructure for scalable, responsible AI
Methodology
Appendix A: Methodology
In this study, Forrester conducted an online survey of 576 IT and business decision-makers of Al and cloud infrastructure strkegy in North America, Europe, and the Nordics to evaluate their organizatirms’ Al infrastructure readiness for Al initiatives Survey parkin.. included decision-makers who are Cdevel executives. vice president. directors, a. senior managers. Questions provided to the participant asked about thee organizations. Al Infratructure, challenges. and future plans. Respondents were offered a small Incentive as a Nank you Nr time spent on the survey. The study began in April 2025 a. was COMpieted in May 2025.
