Wildfires release air pollutants such as aerosols, negatively impacting air quality and weather. Forecasting smoke is fundamental for helping minimize people's exposure, improving weather forecasting, and guiding wildfire-fighting operations. NOAA's Global Systems Laboratory (GSL) is developing a new weather forecast model (the Rapid-Refresh Forecasting System (RRFS)) based on the Finite Volume Cubed-Sphere Limited Area Model (FV3-LAM). RRFS has been tested in real-time at NOAA-GSL and provides experimental weather forecasting at 3km resolution over the contiguous US (CONUS) domain. Here, the smoke emissions, fire plume rise, dry and wet removal capabilities from the operational High-Resolution Rapid Refresh coupled with Smoke (HRRR-Smoke) model are implemented into the RRFS model, hereafter RRFS-Smoke system. Several improvements over HRRR-Smoke include the inline treatment of smoke turbulent mixing within the MYNN PBL scheme and the ingestion of hourly GOES-16/17 ABI satellite fire radiative power data to derive biomass burning emissions, heat fluxes, and fire size. Evaluation of the RRFS-Smoke system focusing on the 2019 FIREX-AQ field campaign and its comparison with the HRRR-Smoke model will be presented.