Energy consumption for lighting constitutes a sizable portion of the overall energy consumption of commercial office buildings. Many smart lighting control products are already available in the market, but their penetration has been limited and even installed systems have had limited use. One of the main reasons is that they tend to control lighting based on universal set-points which are agnostic to the individual preferences of the occupants thus hampering their comfort. The paper will present an automated lighting control framework which dynamically learns the lighting preferences of each user, models his visual comfort and controls the light dimming in a truly personalized manner so as to always control the comfort vs. energy efficiency trade-off. This approach effectively removes the single most important complaint of occupants when using such systems, loss of comfort, and paves the way for their wide scale adoption in order to untap the energy reduction potential of commercial lighting.