Data readiness and knowledge graphs essential for AI success

Preparing data for artificial intelligence requires more than simple access; it necessitates creating an environment where data is accurate, semantically clear, and enriched with business context. Experts emphasize that for generative AI and AI agents to be effective, data must be well-described so that systems can interpret meaning rather than just processing raw values.

OpenText highlighted that while structured data is common in ERP systems, approximately 80% of business information exists as unstructured data, such as meeting notes and proposals. To bridge the gap between digital data and human context, the use of knowledge graphs is recommended. These act as a ‘brain map’ for AI, allowing it to understand relationships, decision-making processes, and organizational context.

Key challenges in this transition include avoiding the creation of fragmented, isolated information bases for AI, which can lead to data inconsistency and high maintenance costs. Instead, organizations should focus on robust metadata and integrated data management to ensure AI can access the most current and relevant information securely.