Health Monitor

As people looking to work in health care, we have seen how patient deterioration doesn’t always happen suddenly, it often begins with small, subtle changes that are easy to overlook until they become critical. This project was inspired by that gap. Instead of relying on fixed thresholds, our system learns what “normal” looks like for an individual patient and detects when their vital signs begin to drift away from that baseline. Using simple, interpretable methods like rolling averages and z-scores, we focused on early awareness rather than perfect prediction. The goal is to provide clear, actionable insights, highlighting not just when something is abnormal, but how and why it changed, so that attention can be directed before a situation escalates. By prioritizing simplicity, transparency, and clinical relevance, this tool demonstrates how even basic AI techniques can support better decision-making in real-world settings.