NYB introduces a new AI drug discovery platform. The platform begins with our proprietary ML model, leveraging protein language models from extensive amino acid sequence libraries and ChEMBL library, NTU's natural product library, and various commercial compounds. This strategy skips the initial need for 3D structural data, accelerating the shortlisting of potential small molecules. These molecules then undergo a comprehensive analysis using Alphafold2 (open source) and Autodock Vina (open source) for molecular docking, and our proprietary Drug-Target Interaction Graph Neural Network (DTIGN), which refines biological activity predictions based on molecule-protein interaction data. This process swiftly yields thousands of hits, which are meticulously assessed for open source ADME-Tox, analyzed through molecular dynamics simulations, and subjected to rigorous preclinical validation. Our streamlined pipeline efficiently narrows these candidates down to a select few promising drug candidates, significantly enhancing the journey from discovery to development and offering a quicker, more efficient route to market for new treatments.