Background <p>Integration of artificial-intelligence based predictive modeling techniques offers exciting possibilities for development and optimization of dosage forms. The Etodolac extended-release tablets were formulated using a Quality by Design (QbD) approach and evaluated through AI-based predictive modeling to optimize polymer ratios and release kinetics for modified release performance.</p> Aim <p>Investigate the potential of machine leaning tools in predicting and optimizing drug release profiles for extended-release tablet.</p> Methods <p>Experimental Factorial design incorporating three input factors like polymer types, concentrations, and diluent concentration at three levels for designing experiments for training models apart from using set of experiments as test set and a validation set to validate model. Further, evaluated the impact on resource utilization, cost effective quality products and comparing different predictive modeling techniques and models to decide the best fit for data obtained from etodolac extended-release tablets.</p> Results <p>AI and machine learning tools enhancing and understanding causes of variations during pharmaceutical manufacturing resulting in control strategies for novel and generic extended-release tablets. Artificial Neural Network (ANN) model presented best fit and predictability of all studied predictive modeling techniques. This was noticeably clear from high R-squared values between the test, validation and training sets generated using the chosen model.</p> Conclusion <p>The study proves the remarkable effectiveness of machine learning tools like ANN as a predictive modeling technique for formulating and optimizing novel and generic extended-release drug products with predefined targets.</p> Graphical Abstract <p></p>