AI-pharmaceutical Intelligence System: A Machine Learning Approach for Symptom-Based Drug Recommendation and Predictive Analytics

The pharmaceutical industry today faces several challenges such as scattered information, lack of price transparency, and limited awareness of alternative medicines. In many situations, patients depend on branded drugs without knowing that similar and more affordable options exist. At the same time, healthcare providers also face difficulties due to changing drug availability and pricing patterns. Most existing digital tools focus only on specific tasks like symptom checking or maintaining drug databases, but they do not provide a complete solution. This paper presents an AI Pharma Intelligence System that combines machine learning, natural language processing, and data analytics into a single platform. The system allows users to enter symptoms in text form and predicts possible therapeutic categories using TF-IDF and Logistic Regression. Based on this prediction, it suggests relevant medicines and also provides alternative options that are more cost-effective. An additional feature of the system is the prediction of drug discontinuation. This module uses factors such as manufacturer frequency, side effects, and market competition to estimate whether a drug may be discontinued in the future. The system is developed using Streamlit and Scikit-learn, making it lightweight and easy to use.