Patient treatment is often guided by snapshots of data, ranging from blood tests to images. However, longitudinal data and corresponding biomarker changes can provide key insights that can be harnessed to modulate patient dosing to optimise efficacy over time. This strategy may even further elucidate changes in combination therapy drug synergy which can be dose- and time-dependent, as well as patient-specific. This talk will address our development of the CURATE.AI workflow, which uses prospectively acquired small data sets to dynamically adjust drug doses. Contrary to conventional approaches, this workflow does not use pre-existing big/population data to treat individuals. Instead, it uses only a patient’s own data to guide only their own treatment. Beyond clinical validation, we will also discuss the role of behavioural sciences, hospital/patient stakeholder engagement, and design towards workflow implementation.