Rationale: The standard diagnostic procedure in drug-resistant epilepsy involves visual inspection of long EEG recordings to localize the seizure onset zone (SOZ). This is a very time-consuming and demanding procedure that might lead to inconclusive interpretations, resulting in erroneous or incomplete diagnosis. To assist neurologists, we have designed BrainFocus, a novel cloud-based software that implements a variety of computational and visualization tools to analyze EEG recordings and provide accurate SOZ localization. Methods: The core technology of BrainFocus is based on a recently developed algorithm that overcomes a major limitation of previous SOZ detectors. Previous biomarkers relied on the detection of preselected spectral activations (e.g., HFOs) and thus fell short to capture the broad diversity of seizure-specific onset patterns observed in clinical practice. In contrast, our algorithm identifies the spectral signatures associated with the emerging seizure-specific onset patterns. To achieve so, the algorithm relies on finding time-frequency windows where two indices targetting locally enhanced oscillations (Global Activation and Activation Entropy) are jointly optimized over brain sites (Vila-Vidal et al., 2020, doi.org/10.1016/j.neuroimage.2019.116410). Then, regions where these patterns are most prominently observed are identified. With a user-friendly interface that has been designed to satisfy clinical needs, BrainFocus enables clinicians to quickly assess multiple-contact EEG recordings, obtaining a report that includes a description of the seizure onset patterns and a list of regions that most likely lie within the SOZ. In addition, BrainFocus features visualization tools that are aimed to organize the information in a clinical-friendly and compact fashion, providing 3D maps of the seizure focus and propagation of the seizure activity across the epileptic network at a glance. BrainFocus is currently developed and tested for intracranial EEG and is now being extended to scalp EEG. Results: In this study, we further validated the method with a larger cohort of patients beyond TLE while assessing the ecological validity of a BrainFocus’ prototype in a realspace environment, For this study, BrainFocus has been applied to ca. 80 seizures from a cohort of 20 pharmacoresistant epileptic patients from Hospital Clínic (Barcelona, Spain), including non-TLE epilepsy. Upon revision of preliminary results, the algorithm achieved high accuracy rates concordant with previous own study (sensitivity and specificity ~95%) (Vila-Vidal et al., 2020, doi.org/10.1016/j.neuroimage.2019.116410). In addition to the algorithm’s performance, testing the software in a clinical setting was proven successfully, speeding up the clinicians’ inspection and analysis, offering intuitive summaries and plots, that helped identify early activation patterns invisible to the human eye. Conclusions: BrainFocus is a novel online software aimed to assist clinicians in the diagnosis of epileptic patients with EEG extracting SOZ relevant information and displaying it in a compact, user-friendly and interactive form. Funding: Please list any funding that was received in support of this abstract.: M.V was supported by a fellowship from ”la Caixa” Foundation, Spain (ID 100010434, fellowship code LCF/BQ/DE17/11600022). A.T.C and M. V. were supported by UPF INNOValora grant, European FEDER funds. G.D. was supported by the Spanish Ministry of Economy and Competitiveness, Spain (grant agreement number PSI2016-75688-P, MINECO/AEI/ FEDER-EU); European Union’s Horizon 2020 FET Flagship Human Brain Project (grant agreement number 785907, HBP SGA2); the Catalan Agency for Management of University and Research Grants, Spain (grant agreement number 2017 SGR 1545).