The Open Data Institute (ODI) and SAP are launching the IDEA (Interchange for Data and Enterprise AI) programme to convert corporate data infrastructure into forms suitable for AI.

The two sides aim to develop open standards through 3 approaches: conducting research, building an independent governance system and running a community.

According to Techzine, the programme starts from common challenges companies face.

Much corporate data is designed for transaction records, reports and compliance and is structured for human use. Such data is difficult to feed directly into AI systems. AI systems that use unprepared data are more likely to produce results with lower reliability, and there could also be compliance issues.

The ODI-SAP partnership has 3 main pillars. First, ODI will build an independent governance model based on its 14 years of experience in public-private partnerships. Second, it will study ways to help chief information officers and chief data officers convert data infrastructure to make it suitable for AI. It will also cover integration with traditional machine learning, generative AI and agent AI, as well as data mesh, data fabric and data products.

Third is a community made up of SAP customers, partners, policymakers and researchers. The community will contribute to developing open standards for enterprise data and share best practices. ODI and SAP also plan to work with external partners to flesh out an enterprise data open blueprint that connects various systems and data sources beyond a single vendor's scope.