This paper proposes an adaptive data-driven novel technique for generating physician's schedules exploring the trade-off between physician waiting time and patient experience. Generally, in hospitals fix time slots are assigned to the patients, without differentiating patients into different categories. Some patients are referral patients who need less time, others are first-time visitors who need more time than referral patients. Some are serious patients who take even more time, so a fixed time slot can increase the waiting time both at the doctor and the patient level. So, there is a need to categorize patients according to their treatment which could help in improving the waiting time of doctors and as well of the patients. This technique trains the patient's data according to their specific category based on the time doctors usually spend on them. This can help in generating a schedule that will minimize the waiting of patients and at the same time improve the hospital performance by scheduling maximum patients ensuring that the same patient's get the best experience. To achieve this a data-driven scheduling algorithm (DDSA) is proposed which uses probability distribution, clustering and doctor's scheduling algorithm to classify patients into different categories. The data is collected through the RFID machines. The patients are given cluster head time for the treatment. Different patient categories are given different treatment times. Then patient and doctor commutative waiting times are calculated. The system applies different set-up times to allow hospital management to trade-off between doctor and patient waiting times. The average waiting time for each patient using DDSA comes out to be seven minutes. Hence, this technique can help hospitals across the globe in improving their performance and patient experience.