The addition of air-entraining agents to wet-mix shotcrete is commonly considered an effective approach to improve its pumpability by creating artificial air voids. Additionally, incorporating fibers into shotcrete mixtures has been observed to increase the air content in fresh concrete. However, there is ongoing debate regarding the impact of the spraying process on the final air content of shotcrete. Some researchers suggest that spraying reduces air content due to the compaction effect, while others argue that the process introduces additional air, potentially increasing porosity and compromising durability. To address these conflicting perspectives, this study investigates the influence of the spraying process on the final distribution of air voids in shotcrete. Samples were collected at various stages: (I) conventionally cast, (II) sprayed without accelerator, and (III) sprayed with accelerator. Additionally, the effect of several parameters was analyzed, including: (A) two surfactant air-entraining agents, (B) accelerator dosage (0%, 5%, and 7%). To assess the air content, specimens sprayed without an accelerator were evaluated using both standard testing methods and micro-computed tomography (μ-CT), while those sprayed with an accelerator were analyzed exclusively through μ-CT. To enhance the accuracy of the analysis, deep learning algorithms were trained to differentiate between cracks and pores, as well as to identify regions of interest. This study aims to provide clearer insights into the effects of the spraying process on air content and its implications for the durability and performance of shotcrete.