What About Time Series AI?

By Rich Staples

With all the buzz around Generative AI, it’s easy to forget that it’s just one type of artificial intelligence. Time Series, Geospatial, Computer Vision, and Graph AI are also growing fast—sometimes even faster than GenAI— particularly in sectors where real-time predictions and pattern recognition are business-critical.

Take Computer Vision: you've seen it in action in manufacturing, retail, and in medical imaging. But Time Series AI? It's not grabbing headlines, yet it's quietly driving massive outcomes behind the scenes.

So What Is Time Series AI?

At its core, Time Series AI is about pattern recognition and forecasting. It identifies patterns in data collected over time to predict future events, improve processes, and reduce risk. If you’ve ever gotten a fraud alert from your credit card company, you've already seen Time Series in action. It's also used to:

Predict equipment failures before they happen
Forecast retail sales and manage inventory
Optimize logistics and transportation
Inform financial models for investing and stock trading

And in healthcare? It’s also made significant progress.

Time Series in Healthcare: Quietly Transforming the Industry

Staffing accounts for up to 55% of a hospital’s operating costs. Predictive models can surface trends and drive efficiency, cutting costs without sacrificing quality.

Here’s a real-world example:
A major hospital system in Texas used Time Series AI to predict ER staffing needs. The model pulled in everything from weather patterns to local events—even who was playing in the Friday night football game. Over time, it evolved and ultimately helped reduce ER staffing costs by 25%.

Another healthcare provider took on the high cost of nurse turnover. By analyzing years of workforce data—think shift patterns, vacation time, tenure, manager changes—they were able to reduce turnover by 20% and increase staff retention in high-stress departments.

These models aren’t built overnight. They require massive data sets—millions of data points across departments, facilities, and time. Especially in critical care, where ICU monitoring happens continuously, Time Series AI helps make sense of chaos, improve outcomes, and reinforce best practices.

One initiative I worked on analyzed antibiotic protocol adherence across departments and hospitals over an eight year period. We found patterns in how protocols were followed (or improved upon!) and used those insights to improve the system. In a clinical environment, it’s not enough to say “follow the protocol”—you’ve got to bake those expectations into workflows, devices, or visible dashboards to make it stick.

Why It Matters Now

Time Series AI is already embedded in many businesses—but, it is not as visible as other types of AI. As AI technology, infrastructure and governance improves, these models are becoming faster, smarter, and more adaptable to dynamic environments. That makes now a great time to reassess where Time Series can make a strategic difference—whether that’s operational efficiency, organizational excellence, or risk reduction.

And if you’re curious how Time Series stacks up against other types of AI: Databricks’ State of Data + AI report (late 2024) does a great job comparing growth and adoption across multiple industries.

Rich Staples is a healthcare technology leader who helps organizations grow by making their customers more successful. He writes about real-world strategies, outcomes, and lessons learned from years leading transformation at companies like Philips Healthcare.