Cognitive Digital Twins (CDTs) are well-established for manufacturing but have had very little implementation in the buildings sector. This paper presents one such implementation: a CDT for a mixed-use (academic/ residential) building developed from the construction BIM. Lessons learned regarding five key elements of CDT development are presented: the integrated data model and supporting ontology; building automation system data acquisition and streaming; data lake; event detection algorithms; and integration and visualization. Insights regarding approach selection, implementation considerations, limitations, and alternatives are presented for each to guide future CDT development.