Tuberculous meningitis (TBM) and Pyogenic meningitis (PM) share similar clinical features, making their differentiation chal- lenging, particularly in the absence of a medical expert. The over- lapping symptoms often lead to diagnostic complexities, potentially delaying crucial interventions. Therefore, it is necessary to develop a system that can differentiate between Tuberculous Meningitis and Pyogenic Meningitis. A Progressively Web-Based Smart System for diagnosing TBM and PM can be built based on laboratory findings by applying Machine Learning (ML) techniques. In the absence of a medical expert, our system aids in the early identification and clas- sification of cases based on established patterns and data-driven in- sights. Alternatively, when a medical expert is available, our system acts as a complementary resource, providing a second opinion and additional diagnostic support. This collaborative approach, combin- ing the expertise of medical professionals with the analytical capabil- ities of our system, aims to enhance diagnostic accuracy and expedite the decision-making process for effective and timely intervention. It is made by bagging three Machine Learning algorithms namely Multi- Layer Perceptron, Support Vector Machine, and Random Forest to forecast a person’s health status utilizing five CSF (Cerebrospinal Fluid) features and also an additional model is used in assessing condition intensity, classifying it as mild, severe, or highly severe meningitis. This combination outperformed various ML techniques with an accuracy of 59.77% when compared to individual techniques like Multi-Layer Perceptron (35%), Random Forest (18%), Gradient Boost (17%), Logistic Regression (20%), C4.5 (19%) and Support Vector Machine (21%) on a real patient dataset collected by private hospitals.