Severe acute kidney injury (AKI) requiring renal replacement therapy (RRT) is a common and serious condition that is associated with poor outcomes. Previous trials and their meta-analysis showed that early initiation of RRT in critically ill patients with AKI was not associated with survival benefit compared to late initiation. Both insufficient delivery of RRT and hemodynamic instability during RRT may be a cause of failure to show the benefit in early initiation of RRT. Particularly, intradialytic hypotension is much common during intermittent hemodialysis in patients with AKI. To reduce intradialytic complications, appropriate prediction, early warning, and immediate management are needed when intermittent hemodialysis is applied to patients with AKI, and artificial intelligence (AI) may make it with convenience. Previously, we developed AI model to predict hemodynamic instability on hemodialysis in real-time, such as hypotension and hypertension. Herein, we could provide information on risk factors; some were well known and others were not. We need more active and validated system to predict, warm, and manage several complications on intermittent hemodialysis, which will alter our practice to apply hemodialysis in patients with AKI.