Most of us, frankly, never think about what happens in the seconds after we dial 911. I spent a lot of years in telecom, so I sort of know the plumbing, and it is still a little humbling to hear how much has changed. That is really why my chat with Zach LaValley, CTO of RapidSOS, stuck with me. His company just picked up a 2026 Stratus Award in the Solution / Platform as a Service (Product) category, and honestly, winning industry awards rarely comes with a story this practical.
A 911 System Built for Landlines
Zach put the history in pretty plain terms. He called 911 "frankly revolutionary when it first came out," and it worked well since everyone had a landline that never moved. In his words, "you knew exactly where someone was standing when they were talking to you on the phone." That is pretty wild to think about now.
The world moved on, of course, and the pipes underneath were a bit slow to follow. Your car actually knows when it crashes. Your phone knows, and your smartwatch does too. Zach put it this way: "We're surrounded by an ecosystem of technology that is aware and able to help us in our times of need." In other words, a 911 call is only one piece of a much bigger picture, or as he said, "It is the call, but it's more."
The scale is honestly hard to overstate. The 2021 911 Profile Database Report counted 213,652,929 calls across 45 states, and roughly 89 percent of them came from wireless phones, so the old landline picture really is history. Those calls landed at 4,637 primary answering points, which is quite a patchwork. Zach was pretty candid about why change is slow, since there is no national 911 and, as he said, "you kind of have to make this change 6,000 times across the United States."
Adding a Lane Instead of Ripping Out the Road
Now, here is where Zach got interesting. RapidSOS could have shown up with its own hardware and its own network, and plenty of vendors do exactly that. He very deliberately passed on it. His reasoning was actually simple: "When you go in and you bring your own stuff, you create another silo." Fire, EMS and police then arrive with different pictures of the same emergency, and that is just a bad outcome for everybody.
Trust is the real currency in public safety, so I asked how a vendor earns it with agencies that have run on the same systems for decades. His answer was a highway metaphor I really loved: "we're trying to add a lane and that lane is fast." Agencies can send a little traffic down it, get comfortable, and shift more over time, since the old road is still there if anything goes sideways. Or, as he put it, "you don't want to just rip it out under a promise that the new one is better."
He walked me through three ways that people and AI can share the work. A person can stay fully in the loop, with the AI suggesting the next question and the dispatcher doing all of the talking. A supervisor can sit on the loop, just watching a dashboard of AI-handled non-emergency calls and stepping in when something looks off. And a QA manager can sit over the loop, which is a really big deal, since a human team might sample about 1 percent of calls and an AI can review all of them. His summary was tidy: "You just have to pick your use case and pick the right human AI relationship."
Where the Old Way Drops Information
Zach explained what happens when a connected device is not part of a network like his. Honestly, it is often a game of phone tag. A home security system typically fires off, a monitoring center calls the homeowner to check, and if it is real, somebody relays everything to 911 by voice. He said 99 percent of those alarms are not real, which tells you how noisy those signals can be. On the network, he said, "all of that literal game of telephone goes away," and the data lands straight in the systems 911 already uses.
He shared a story from Cass County, Michigan, that really stuck with me. A crash report actually showed a car off a county road, nobody called, and the driver was unconscious. According to Zach, dispatchers had it within 13 seconds, responders arrived in about four minutes, and they pulled a man out of a burning car, which is pretty remarkable.
Speed like that matters most where help is farthest away. According to NHTSA research, EMS reached rural crash scenes within 10 minutes only about 55 percent of the time, versus roughly 88 percent in urban areas, which is quite a gap. Those numbers come from 1994 through 2003 data, so they are a bit old, but the gap still tells the story.
Check Your Phone, Then Check Your Car
Zach closed with a quick plug that I think everyone should act on. Go into your phone's emergency health profile and enable it so it can be shared during a 911 call, and he stressed that it only gets shared during a 911 call. AARP has a pretty simple walkthrough for setting up a Medical ID on your smartphone. He suggested, as a sort of bonus tip, asking your tech vendors, car makers included, whether your devices are connected and through whom.
That sort of practical, human-centered thinking is exactly what the crowd-sourced judging behind our Stratus Awards tends to surface. For other teams eyeing winning industry awards, the RapidSOS story is a pretty decent template: solve a real problem, earn trust slowly, and be able to explain it in plain language.
So, want the whole conversation? Press play on the video above, and subscribe to the Winners' Circle podcast so you never miss a winner. Know a team doing work like this? You can nominate them for the Stratus Award for Cloud Computing.