Disclosed are systems and methods for proactive service health detection. An example method may include receiving, by a first computing system, first data associated with a first application at a first time and second data associated with the first application at a second time. The example method may also include receiving an indication of an anomaly associated with the first application at the second time. The example method may also include comparing, by a first natural language processing model on the first computing system, the second data and the first data. The example method may also include determining, by the first natural language processing model and based on the comparison, a first difference indicative that a portion of the first data is different than a same portion of the second data. The example method may also include receiving third data associated with the first application. The example method may also include determining, by the first natural language processing model, that the first difference exists in the third data. The example method may also include automatically initiating a first action to prevent or mitigate a second anomaly associated with the first application.