We have demonstrated the first ever example of closed-loop feedback control over plasma drug levels informed by real-time in-vivo drug measurements. We achieved this using a sensing architecture that is independent of the chemical reactivity of the target molecule, rendering the approach, at least in theory, generalizable to any drug. Here we propose a research and development program aimed at translating this innovative breakthrough into a powerful new tool applicable to both pharmacological research and clinical practice. By improving (by orders of magnitude) the precision with which we can control in-vivo drug levels, the proposed technology would render pharmacokinetics an experimentally controllable parameter, allowing researchers to more precisely map the relationships between pharmacokinetics and clinical outcomes, to eliminate pharmacokinetic inconsistency as a confounding experimental variable, and to accurately simulate human pharmacokinetics in animal models. Ultimately, such a technology could be used in the clinic where, by optimize dosing moment-to-moment in response to a patient's changing metabolism, it would ensure that a therapy's benefits are maximized while its risks are minimized.