HIV management faces challenges in utilizing viral load data for personalized treatment. To overcome this, we developed a groundbreaking prescriptive model that leverages big data analytics and personalized recommendations to optimize HIV care. Our objectives were to train a prescriptive model that tracks HIV viral load results for patients by predicting viral load trends and generating personalized treatment recommendations based on the tracked viral load results. This model combines cutting-edge techniques, including reinforcement learning. By analyzing historical viral load data, treatment regimens, adherence information, and patient characteristics, the model predicts viral load trends and identifies at-risk patients. Furthermore, Natural Language Processing (NLP) extracts crucial information from clinical notes, enriching patient profiles for personalized treatment recommendations. Integrated with big data tools, the model efficiently processes large volumes of data, providing real-time insights for informed decisions. The research methodology followed the CRISP-DM model, encompassing business understanding, data preparation, modeling, evaluation, and deployment. An artificial intelligence (AI) assistant for clinical decision support through self-supervised pretraining and continual learning was successfully deployed. The random forest (RF) algorithm was the most effective in this model, giving an accuracy of 81%, a recall of 90%, and an F1 score of 90%. The outcomes were improved medication adherence through personalized recommendations and optimized treatment plans based on viral load predictions. This model contributes to achieving the 95–95-95 UNAIDS target for viral load suppression. The model makes a significant contribution to the fight against HIV/AIDS by harnessing advanced algorithms and big data, revolutionizing the management and improving outcomes for all patients.