This paper summarizes the ongoing research on the Bishopsgate Tower in the City of London designed by Kohn Pedersen Fox Associates. We present a pre-rational geometry computational solution targeting a constraint-aware exploration of the architectural design-space, while interactively optimizing building performance in terms of constructability and cost-efficiency. We document a novel approach in building metrics optimization supported by parametric technologies and embedded analytical algorithms. The process is indicative of how computational methods will develop in the future and help designers find solutions for increasingly complex spaces.