The 2023 fire season was a record-breaking natural disaster event in Canada with more than 15 million hectares (Mha) of forests consumed by wildfires. Smoke from wildfires resulted in extremely poor air quality across Canada and impacted cities in eastern USA. Smoke plumes were also transported long-distance across the Atlantic impacting cities in Europe. The FireWork air quality forecast modelling system operated by Environment and Climate Change Canada (ECCC) has been demonstrated to be a valuable tool during these extreme smoke episodes. The system captured the timing and duration of the pollution, and adequately forecasted the PM2.5 concentrations in many regions across the country. A new research version of the system has fire emissions modelled within the GEM-MACH chemistry transport model with coupled meteorology feedback. The system accounts for aerosol direct and indirect effects on radiative transfer and cloud microphysics. In regions impacted by smoke, the fully coupled model simulations showed reduced surface air temperatures, reducing forecast biases relative to observations. In this work, we will present operational analysis of the FireWork system for the 2023 fire seasons, at the same time, show research application of the coupled model on the study of wildfire smoke and aerosol effects on regional weather.