At present, diesel engines are confronted with the formidable task of reconciling aggressive emission regulations with the preservation of intended engine performance. In order to tackle this concern, India can implement a comprehensive strategy that includes waste reduction, recycling programs, and the innovative use of waste-to-energy technologies. The potential synergistic effects of combining fuel mixtures with cerium oxide nanoparticles in CRDI diesel engines remain unexplored. This study employs response surface methodology and MLP regression to investigate these effects. This study explores the experimental design includes variation in blend ratios of karanja biodiesel (upto 20 %), compression ratios (18–22), and injection pressure (400–1200 bar), as well as CeO2 nanoparticle concentrations (50–150 ppm) using 10 % biobutanol. A combination of response surface methodology (RSM) and artificial neural network (ANN) techniques is used to assess and optimise the influence of diverse process parameters on engine responses. The machine learning processor (MLP) regressor model is more capable of forecasting engine responses than RSM modelling, as it has lower MAE, RMSE, and higher R2 values. According to the study, the best blend ratio, CR, Injection Pressure, and CeO2 concentration for the best performance and lowest emissions are 9.97 % biodiesel blend, CR of 20, an injection pressure of 797.4 bar, and CeO2 nanoparticle concentration of 100.1 ppm. The MAE value of the MLP regressor-based model ranged from 0.0011 to 0.0119. In terms of RMSE, MLP regressor-based model outputs are within reasonable boundaries. This study highlights the optimization of engine operating parameters and the application of nanotechnology in biodiesel blends. These findings can lead to the development of durable and eco-friendly fuel alternatives for internal combustion engines, promoting sustainable transportation.