Life Cycle Assessment (LCA) is an essential tool for quantifying the environmental burdens of products and processes, critical for advancing sustainability goals. Central to the effectiveness of LCA is the Life Cycle Inventory (LCI) phase, which requires reliable data to reflect the environmental footprint of products accurately. However, LCA practitioners often encounter data gaps that can compromise the assessment’s accuracy. To address this, we explore the integration of Machine Learning (ML) to enhance LCA data quality, particularly in the LCI stages B to D, which focus on product use, end-of-life, and beyond-life phases. This chapter introduces a novel framework that leverages ML to overcome LCI data challenges, emphasizing reducing the embodied carbon of construction products. We extract existing data from the Environment Product Declaration online library and apply natural language processing to interpret this unstructured data. Subsequently, we employ a random forest algorithm, a robust ensemble tree-based ML method, to refine the data analysis. We present a pilot study that validates the feasibility of our ML-enhanced framework. The incorporation of ML addresses the voluminous data in LCA. It augments the analytical capacity, thereby improving the precision and reliability of both LCI and Life Cycle Impact Assessment (LCIA) datasets. Consequently, our approach yields higher quality LCA outcomes, offering a more reliable basis for environmental impact evaluation. In summary, the successful application of ML in this research bridges the critical data gap in LCI for construction products, paving the way for a more sustainable industry through improved accuracy in environmental impact assessments and more informed decision-making in green product innovation.