Sepsis is a severe medical complication stemming from an infection in the body and can result in tissue damage, organ failure, or even death. About one in three cases result in death. Sepsis is a disease that can be seen in neonatal, pediatric, or adults. Early detection helps reduce the mortality rate as a reduction in delay in antibiotics therapy. Several studies have demonstrated that treating sepsis patients quickly increases their likelihood of surviving. Researchers have carried out work concerning detecting sepsis using blood cultures and biomarkers. This research presents a comprehensive exploration of machine learning algorithms, specifically the K-Nearest Neighbors (KNN), Nearest Centroid Classifier (NCC), and Support Vector Machine (SVM), for sepsis detection. The KNN model emerges as a standout performer, achieving a remarkable 98.03% accuracy on the test set, underscoring its proficiency in accurately classifying instances. However, a discernible trade-off between precision (48.39%) and recall (13.27%) prompts the recognition of the necessity for a balanced approach to detecting positive instances while minimizing false negatives. The NCC and SVM models exhibit accuracies of 82% and 74%, respectively, with diverse precision, recall, and F1-score values. The study contributes to the ongoing advancement of machine learning techniques in sepsis detection, offering valuable insights into algorithmic performances.