Abstract Atrial fibrillation (AF) is the most common arrhythmia especially in the aging population. It is associated with increased risk of mortality and morbidity. At the present time, management of AF has focused on risk factor modification, rate or rhythm control and anticoagulation. Evolution of clinical trials in the management of AF have revealed that ablation seems superior in reducing the burden of AF and controlling the symptoms compared to pharmaceutical agents. However, the benefit of ablation decreases over time and patients frequently require more than one ablation. Earlier ablation in the course of the disease is more beneficial as failure of therapy is related to duration of AF and size of left atrium. After two decades of investigations with varying methods of ablation, we have only marginally improved the clinical outcome. The ablation procedure is time consuming and only a fraction of patients are undergoing this procedure. A robust criterion of prediction of successful ablation will be beneficial for patient selection and maximize the utilization of invasive therapies. With this highly collaborative and multiscale study, our long-term goal is to develop effective models and discover factors that indicate severity of AF that can be helpful as therapeutic targets and to predict prognosis. Our objective is to identify patients who have increased risk of recurrence after ablation for AF by taking advantage of the intracardiac electrograms from left atrial map and inflammatory biomarkers from blood samples obtained pre-procedure. The central hypothesis is that domain-specific machine learning/ artificial intelligence algorithms derived from multimodal data can predict the type of AF, severity of AF as indicated by abnormal areas in the left atrium and clinical outcomes of AF ablation. To directly test the hypothesis, we will enroll prospective consecutive consenting patients who present for AF ablation therapy. Pre-and post-ablation left atrial map and blood samples drawn for biomarker analyses will be used for study purposes.