Epilepsy is a neurological disorder marked by recurrent seizures that afflicts an estimated 60 million people worldwide. One-third of patients cannot control their seizures using medication, and responsive neurostimulation (RNS) therapy provides a treatment alternative when surgical removal of the seizure onset zone is not an option. The RNS System is a neuromodulation therapy that uses an implantable device to continuously monitor neural activity and respond with electrical stimulation pulses when abnormal activity is detected. Most patients see a significant reduction in seizure frequency with RNS therapy, but seizure freedom is rare. The mechanism of RNS action remains poorly understood and no validated clinical biomarkers exist to indicate whether a patient will be a good RNS candidate, where electrodes should be implanted, or what stimulation parameters to use to achieve the best outcome. To address this gap, I leverage quantitative tools from network theory to develop models and biomarkers that may improve RNS therapy. First, I formulate a novel, patient-specific, linear control model that captures how the time-evolving dependencies between brain regions dictate the impact of exogenous stimulation. The model allows me to probe the energetic requirements of external control at various brain regions throughout the onset, propagation, and termination phases of seizures. Next, I pioneer a network-based biomarker derived from intracranial EEG (iEEG) data to differentiate those who eventually respond to RNS therapy from those who do not. I validate my measure using a multicenter dataset built using a federated approach. Finally, I complete a normative modeling study with short-term iEEG recordings from the epilepsy monitoring unit and longitudinal device-recorded iEEG to assess strategies for RNS electrode placement. The results of this thesis demonstrate how network models can be used to assess potential RNS control strategies, recommend candidates for RNS therapy, and advance the goal of using a quantitative framework to improve clinical guidance on neurostimulation therapy in epilepsy.