As organizations across industries rush to harness the transformative power of AI, a key challenge has emerged: Most enterprise data was created for human use—such as transactions, compliance, or reporting—but not for AI. Applying AI data infrastructure systems on unprepared data can lead to frequent failures, compliance risks, and unreliable results, highlighting the need for proper data infrastructure.
To provide a solution to address these needs, the Open Data Institute (ODI) has partnered with SAP to develop and manage a program of research, community building and peer learning. This program is designed to help businesses of all sizes prepare their data infrastructure for AI, laying the foundation for reliable and scalable adoption. A well-designed data infrastructure will be key to ensuring the success of these initiatives.
Sustainable evolution of data infrastructure
The collaboration between both entities will directly address this gap, bringing together business actors, industry partners and academic experts to jointly design interoperable standards and practical frameworks. The goal is to enable organizations to build robust data infrastructure and databases fit for the AI era, thereby facilitating a sustainable evolution of their data infrastructure.
The ODI and SAP will work along three main lines:
• The first will establish a solid and independent governance model for the program, building on ODI’s 14 years of experience in managing multi-stakeholder initiatives in the public and private sectors, thus reinforcing trust in the data infrastructure generated.
• The second line will develop rigorous research to guide CIOs (Chief Information Officers) and CDOs (Chief Data Officers) on how to prepare business data for AI. The intersection between traditional machine learning, generative AI and agent-based AI will be analyzed, along with different management approaches such as data lakes, data mesh, data fabric and data products, all of which are essential for a modern data infrastructure.
• The third will build and activate a diverse community of SAP customers, partners, policy makers and academics to share best practices, define research priorities and collectively advance open standards. This collaboration will drive the development of an AI-ready data infrastructure aligned with market needs.
Louise Burke, CEO of the ODI, said: “Over the next decade, AI will define business competitiveness, but competitive advantage does not come from AI models alone. It comes from the quality, governance and autonomy of the data that underpins them. Most organizations have data that is simply not AI-ready, and the consequences of not addressing this correctly – ranging from biased results to regulatory non-compliance – are significant.
Accessible AI for all companies
Through this collaboration, ODI and SAP combine the expertise, research and community needed to provide organizations with the action plan they need. Our goal is to make AI-ready data infrastructure accessible to businesses of all types and sizes, based on open standards that no vendor exclusively controls.”
For his part, Irfan Khan, SAP Chief Product Officer for Data & Analytics, said: “As companies expand the use of AI in 2026, the real gap will be trust in data, not technology. Organizations with regulated and integrated data will have a much better chance of achieving better performance and measurable results. A critical step is establishing a business data fabric (Business Data Fabric) that ensures that AI agents have the context necessary to understand the business and its impact, thus consolidating an effective data infrastructure.”
During the development of the project, ODI and SAP will create a research program committee to manage and guide priorities, publish results and organize events with industry and academic experts, promoting the exchange of knowledge around data infrastructure.
Organizations with regulated and integrated data will be much more likely to achieve better performance and measurable results.
The alliance is actively seeking the participation and financing of actors from the sector and other areas, in order to work together on the design of an open and interoperable model. This model will allow us to evolve towards a truly AI-ready enterprise data infrastructure.
