In this study we have developed a computational end-to-end framework to investigate properties of organic corrosion inhibitors responsible for inhibition of mild steel in HCl solution. Several studies in the past have reported Quantitative Structure-Activity Relationships (QSAR) based models for predicting corrosion inhibition performance for steels. However one of the major limitations in these studies is that the authors have restricted themselves to use of only a single class of molecules. Using advanced machine learning algorithms such as support vector machines (SVM) random forest (RF) etc. we have developed a robust computational predictive model of corrosion inhibitors which is not limited to a particular class of molecules. Our model is based on quantum chemical and molecular structural parameters. Our data visualization frameworkprovides users much deeper fundamental understanding of the effect of each independent variable on corrosion inhibition. The study has identified features that have higher and consistent impact on experimental inhibition efficiency. Using these parameters we have also designed novel molecules which are having higher inhibition efficiency as predicted by our model. Our model will also help in discovering and screening of novel corrosion inhibitors which can replace existing toxic inhibitors.