Ortecha’s Post

Gartner CIO Symposium in Barcelona with my friends at Ardoq, last week was phenomenal. Fascinating thought provoking talks from some of the brightest minds in the industry. Loads to unpack. I’ll be sorting a free webinar to cover the key themes seen across the industry but for now here are 3 that really stood out. For those who are members of the EDM Council’s AI SIG these will be discussed in more detail as part of tomorrow’s meeting. 🎓 Neuro-Symbolic AI & Digital Twins Increasing focus across industry of multi-modal AI combining the creativity of LLMs (statistical AI) with the fact based reasoning of graph (symbolic AI). A combination of vector embedding with the context of semantics enables grounding of responses and the ability to check validity. While this is not new to those who have been following Tony Seale, Juan Sequeda Dean Allemang or OriginTrail for any time, the concepts are becoming more mainstream and accepted as the best way forward. This also necessitates a far greater commitment to semantic modelling and the modelling of enterprises as Digital Twins (which of course Ardoq is perfect for capturing and modelling an organisation’s Enterprise Architecture). 🎓 Agentic AI Juan Sequeda and Dean Allemang will be the first to remind us that agents are nothing new, increasingly the use of Generative AI approaches are gravitating towards augmenting and delegating human activities as “agents”. This means they need to be managed in much the same way as human resources over time and may have impacts on geo-political factors such as legislation, tax, visas, ethics and sovereignty when delegating work (and even travel) traditionally reserved for humans. 🎓 Accelerated focus of “Knowledge Management” and TRiSM Incidentally Agentic AI necessitates the need for high quality information (we all know this Ben we haven’t been hiding under a rock!) BUT the real pivot is towards unlocking all the knowledge in unstructured formats across your enterprise (what was sometimes referred to as a separate discipline “knowledge management”) is now a central part of data management practice if it is to stay relevant in this new age of AI. This necessitates new paradigms on locating, keeping up to date and making unstructured knowledge available. The concept of TRiSM (Trust Risk & Security Management is going to need to extend and improve access controls so agents can access the necessary information they require and execute tasks we need. (The irony as the great "under-served masses" of humans in organisations rarely have access to or knowledge of all the information they need and only now we as organisation are giving it the due attention it deserves only to organise it better of AI to help us.) Ping me for more. I'll be arranging briefings on all the insights gleaned from a season of conferences (IRM UK, DataVision, Gartner, DKGCon) and more for our clients, Ortecha alumni and Ortecha partners in December. #AI #datamanagement #knowledgegraph #semantics