Several workplace technologies have already surpassed 80% adoption rates for operational functions; predictive maintenance, data warehousing, energy and emissions management, demonstrating that the infrastructure layer is maturing. The strategic intelligence layer is what remains underdeveloped. And the organisations that are closest to having it are not necessarily those that spent the most on technology, they are those that made deliberate decisions about data ownership, integration architecture and operating model design before they bought anything. JLL

Which brings me to the question I want this community to answer from lived experience - How should your PropTech capability be structured for tomorrow's AI-native system — in-house, outsourced, or hybrid?

Specifically:

— If you outsource your CPIP/IWMS to an IFM provider or managed service, do you retain meaningful control of the data model and integration architecture — or does the AI-native capability you'll need in five years effectively belong to your supplier?

— If you are building in-house, where have you drawn the line between what you own and what you buy — and how are you keeping pace with a vendor market moving faster than any internal team can match?

— Has anyone in this community piloted a genuinely agentic AI workflow across the portfolio-to-projects pipeline — not a chatbot on top of a legacy system, but something that reasons across connected datasets and surfaces decisions? What does that actually look like in practice?

— And for those mid-transition: what would you tell your past self about the sequencing? Data first, then platform, then AI — or did you find a different order that worked?

The blueprint for the AI-native CRE system is being written right now, partly by vendors, partly by the occupier organisations willing to architect it themselves. Interested to understand where peers in this community are in that journey and whether anyone is further along than the rest of us suspect.