Predicting the Unpredictable: Why AI Is the Next Frontier in EMS

After October 7th, we faced a sudden, painful truth: nearly 1,000 of our trained EMS volunteers were drafted into the IDF overnight. That left entire communities without their local first responders.

We had a choice, either spend 6–9 months training replacements, or ask a harder question:

Could we use 18 years of emergency call data to predict where the next call will be?

Turns out, we could.

Built in Crisis. Designed for Speed.

We brought in former Unit 8200 engineers and United Hatzalah’s own tech team, and sat them down with our dispatchers and medics. No theory. No sales pitch. Just one mission:

Help us hit the 90-second response time, even with fewer boots on the ground.

Within weeks, we had a working model. Today, in major Israeli cities, we can predict, with over 90% accuracy, where the next emergency will happen within a two-minute radius.

This isn’t science fiction. It’s field-ready AI, saving real lives.

Why It Works

With this AI tool, our ambulances and ambucycles don’t wait for a call. They hover where the call is likely to happen, based on traffic, time of day, local health history, and hundreds of other signals.

Volunteers don’t waste their shift circling blocks. They go to the zone. They wait. They get the call. And they’re already half a block away.

Less time on the road. Fewer vehicle crashes. More lives saved.

What’s Next, And Why It Matters to You

In 2006, we introduced location-based volunteer dispatch, before iPhones even existed. Everyone told us it was too ambitious. Today, it’s standard in dozens of EMS systems around the world.

I believe AI-powered positioning will be next.

If you run an EMS agency, work in public safety, or invest in emergency tech, you should see this model. Not because it’s flashy. Because it works.

Invite Me to Speak

I’ve briefed ministries, militaries, and medics around the world, from war zones to boardrooms. If your conference, summit, or university is exploring the future of EMS, I’d be honored to share what we’ve learned (and what’s coming next).

➡️ Invite me to speak: Let’s talk about field-tested AI, not slideware.📩 Speaker inquiries – Contact here