Globally, 186 million individuals live with infertility, which causes emotional, financial, social, and physical distress. In vitro fertilisation (IVF) as effective treatment has been increasingly utilised during the last decade to help with infertility. However, despite significant advances, the success rate is relatively low; about 30% of embryo transfer leads to a pregnancy. One reason for this low success rate is the lack of an efficient method to select a high-quality embryo with a high probability of becoming a baby. Traditionally, embryo quality has been assessed by morphological grading and scoring, which is very subjective, inaccurate, and inconsistent. Several studies tried to introduce new embryo assessment strategies. Embryo monitoring by microscopy proved helpful. However, it is expensive and laborious, making it challenging in routine laboratory settings. Invasive preimplantation genetic testing for aneuploidy did not improve the IVF success rate. Non-invasive biochemical methods, including microRNAs and metabolic activity measurements in embryo-spent media, were unreliable and inaccurate. One possible explanation is the challenges of measuring molecules in the minimal quantity of human embryo materials. Also, the acquired molecular data is noisy, making it difficult to be analysed using traditional bioinformatic tools. We have developed a highly sensitive technology to profile all cell-free RNA types secreted out of embryos (small RNA and mRNA). Also, we use artificial intelligence (AI) techniques to analyse RNA data to predict the likelihood of pregnancy. This research aims to improve the identification of high-quality embryos non-invasively by combining our innovative molecular technology and AI. This methodology will be accurate and reliable for the clinical routine, leading to a paradigm shift in modern IVF practice.