Accurate and timely disease diagnosis is a cornerstone of effective medical treatment, yet it remains a complex and challenging task, particularly for less experienced healthcare professionals. This paper introduces a novel system for disease diagnosis by integrating clinical symptoms with test results, aiming to support healthcare providers in making more informed and reliable decisions. The dataset used in this study comprises 66,508 records collected from hospital’s software in healthcare facilities, which includes both clinical symptom data and laboratory test results. A neural network model is proposed to process and combine these multi-modal inputs. Our model obtains a high accuracy of \(74.49\%\) and an F1 score of \(65\%\) when tested on 13, 302 data entries. Additionally, we implement a Web Service API in real-world applications, offering real-time diagnostic assistance and enhancing the overall efficiency of healthcare delivery. This approach contributes to more effective patient care by improving diagnostic accuracy and decision-making in clinical settings.