Objective: Subclinical mastitis (SCM) in cows is a major challenge in dairying not only in disease management but also in financial issues. The objective of this study is to predict an innovative and sustainable approach for the identification of bovine sub-clinical mastitis using machine learning techniques targetting milk biomarkers like electric conductivity and total dissolved solids. Materials and Methods: The field data on milk electric conductivity (EC) and total dissolved solids (TDS) will be assimulated and connected to a central network system for cross-matching with the library database to predict the result. The cut-off value of milk EC and TDS as standerdize previously would be the exploratory data in machine learning and the output of which need to be translated into language by artificial intelegnce. Results: The optimal EC cutoff value for SCM detection in dairy cows was standardize previously as 6159 μS/cm or 6.16 mS/cm and TDS as 3100 mg/l of milk by examining milk from 108 suspected animals. An automated machine learning based message (study in progress) is assuming to notify the onset of SCM and thus deliver the information to the veterinary/ regulatory authority or dairy owners to predict the severity of SCM and taking necessary actions. Conclusion: This model is assuming to serve the one-step veterinary diagnostic service for bovine subclinical mastitis in Bangladesh.