Plastic pollution is a global environmental crisis, with millions of tons of plastic waste accumulating in landfills and natural ecosystems. Widely used poly(ethylene terephthalate) and durable nylon are among the most present synthetic polymers in our lives, but their persistence in the environment poses significant challenges to waste management and sustainability. Traditional recycling methods, including mechanical and chemical recycling, are often inefficient, energy-consuming, and limited in scalability. Enzymatic degradation presents a promising alternative by enabling the breakdown of plastics into their monomeric components under mild conditions, allowing for potential recycling and circular economy applications. Several PETase and nylonase enzymes have been identified, notably IsPETase from Ideonella sakaiensis and various bacterial and fungal nylonases. However, natural PETases and nylonases exhibit significant limitations, including low catalytic efficiency, instability at high temperatures, and substrate specificity constraints. Industrial applications require enzymes that function efficiently under harsh conditions, such as high temperatures, varying pH levels, and the presence of contaminants. This research aims to address these challenges by leveraging generative AI models for the de novo design and optimization of PETase and nylonase enzymes. Using machine learning-driven protein engineering, structural bioinformatics, and after experimental validation, this project seeks to develop highly efficient enzymes for plastic biodegradation.