The Normalized Hot Spot Indices (NHI) is a multi-channel algorithm developed to map thermal anomalies through the Multispectral Instrument (MSI) and the Operational Land Imager (OLI). The algorithm runs operationally under the Google Earth Engine (GEE) platform, allowing for the analysis of volcanic thermal features (e.g., lava flows/lakes), through plots of hot pixel number, total SWIR radiance and hotspot area. In this work, we present the automated module of this tool, named NHI system. The latter provides automated notifications about volcanic thermal anomalies, at the global scale, detected over the past 48 hours, whenever the NHI web site (https://sites.google.com/view/nhi-tool) is accessed. Results of the first six months of operation are assessed through the analysis of satellite imagery, and the comparison with some well-established programs for the global volcano monitoring. The low false positive rate (around 15%, including vegetation fires and data issues), and the successful identification of small high-temperature features, show that the NHI system may successfully integrate information from high-temporal/low-spatial resolution satellite data, despite some limitations (e.g., temporal sampling of the combined Sentinel-2 and Landsat-8 observations; delay of data ingestion in GEE). The recent ingestion of Landsat-9 data within the system has further extended performance of the NHI system in supporting the surveillance of active volcanoes from space.