When the legacy non-emergency medical transportation (NEMT) companies were founded 30 years ago, it was a very different technological world: Brokers relied on fax machines and on-premises software, and systems were fragmented from one region and team to another. Health plans didn’t really have a holistic view of how their members’ transportation benefits were used or administered because the data either wasn’t collected or lagged far behind rides.  

In some cases, that's still the way it's done today, and the costs are real: A study using Medicaid claims data found that reliable transportation to dialysis appointments saves Medicaid more than $41,000 per patient annually in avoided complications and hospitalizations. The difference often comes down to the quality of the platform behind the benefit.

When we created SafeRide Health in 2016, we were inspired by companies such as Uber and Netflix that were disrupting longstanding industries using new technology. We believed the non-emergency medical transportation space was primed for just such disruption.  

SafeRide anchored its operations in a proprietary technology platform that unifies the entire transportation ecosystem into a single, cohesive environment. Rather than relying on disconnected systems for scheduling, dispatch, routing, GPS tracking, and reporting, SafeRide brings all these functions into a single secure, cloud-based architecture. This allows for consistent benefit administration across markets, transparent oversight, and a higher-quality, more predictable experience for our Medicare Advantage and Medicaid clients and their members.  

Today, legacy NEMT brokers are catching up and building out their tech infrastructure, while we’re focused on making our processes, tools, and operations more efficient. A lot of that is through automation and AI. We’ve given health plan members the tools to schedule their own rides, and we’re using AI and machine learning to maximize the use of the data we collect, making ride dispatch and routing as efficient and accurate as possible. From the beginning, we believed we could improve healthcare transportation while also preventing fraud, waste, and abuse that can drive up costs and bog down the system. Here are a few ways we’re doing it, and what it means for health plans managing costs, member satisfaction, and benefit integrity.

Using AI to Make Humans’ Jobs Easier and More Effective

Humans are so good at many things, but when you give them targeted AI tools, they can become even better at their jobs. For instance, we’ve built an enhanced Interactive Voice Response (IVR) system and AI agent that fields phone calls and directs members to the best solution, often saving our call representatives bandwidth for critical needs while ensuring that members continue to get fast, accurate, human-centered service.  

When a member calls SafeRide, our agents and the IVR have secure access to their past rides, their health plan's benefit structure, their eligibility, and their specific mobility needs. Many of those calls will be referred to agents in our call center, and we’ve given those agents AI tools that make their jobs easier and calls shorter and more efficient. All calls are recorded for sentiment analysis and quality control. About a quarter of ride bookings are handled quickly and seamlessly by the IVR, freeing our care team to focus on the calls that genuinely require human judgment, empathy, and problem-solving.

For members who prefer to schedule rides on their own, we’ve created the MySafeRide mobile app on iOS and Android, as well as a web portal, where they can book rides and see the current state of their transportation benefit. We’re focused on tactical initiatives that effectively leverage the best of both worlds to make transportation even more seamless for the member.

Using AI to Make the Most of Our Vast Data

SafeRide collects multiple data points for every ride: was it on time, who was the provider, how quickly were calls answered, how much did the ride cost, did they use our app, what kind of vehicle was used, did they need assistance from the driver, how did the member feel about the ride, etc. All of that flows into our data lake along with data from third-party systems like Zendesk and Salesforce, which store information about complaints and grievances, provider compliance with licensing and credentialing requirements, and more. We aggregate these data points and use them in real-time to monitor and improve our operations. Then, on the back end, once all services are complete, we provide our clients with very granular details through regular and ad hoc reports.  

We're now working on making this data work even harder, restructuring it so AI can access the information in a more efficient manner, quickly getting to the heart of why a complaint or grievance happened, for example, or tracking general trends that we see on the supplier side and what’s driving them.  

Routing Algorithms and Improved Address Accuracy

At SafeRide, routing and trip assignments are determined by a combination of algorithmic optimization and real-time operational judgment. Our Ride Assignment Algorithm powers about 95% of all rides and takes many factors into account when dispatching a ride to a member, including supply and demand, required modality, and historical provider performance. Leading NEMT providers and rideshare options are scored based on all these factors, with the highest-scoring provider scheduled for the ride. These calculations all take place in seconds. If a transportation provider cannot fulfill a trip, the ride automatically reenters the queue and is reassigned to the next best option, so the member doesn’t miss a ride or their appointment.  

Once a ride is underway, SafeRide uses GPS tracking to maintain continuous visibility. Trips are monitored through our proprietary driver app or through ATMS integrations used by many NEMT fleets. This means SafeRide can validate pickup and drop-off locations, make sure rides are on time, identify unexpected delays, and intervene immediately when a member needs assistance. Members and caregivers see the same information in the MySafeRide app, giving them reassurance about where their vehicle is and when it will arrive.

We’re also using AI to validate address accuracy, with tools that can analyze a given address and make any necessary corrections to resolve variations. For example, a member might have most of their provider’s address right, but in a different format than the official map, or missing part of the street name (i.e. Lake Ave. v. Lake Ave. N..). The AI can pull together multiple address and provider sources to correct those small inconsistencies, ensuring the ride is approved and that they get to the correct location. The member has no idea this is happening behind the scenes, but it makes the interaction much smoother and more accurate.  

Technology-Enabled Fraud, Waste, and Abuse Controls

SafeRide already has multiple guardrails in place to detect and prevent fraud, waste, and abuse. Our benefit integrity framework is extensive: It validates benefits and eligibility in real-time at booking; automatically enforces rules related to modality, distance, and authorizations; uses GPS to validate trips and routes; detects duplicate rides, inflated mileage, and utilization anomalies; and creates full audit trails and role-based access controls.  

Now we’re working to use all that information to assign a probability score to each ride, so instead of being reactive, we can be even more proactive in preventing FWA. For scheduled rides that have a high risk of being wasteful or fraudulent, we can have our specialized team review them in greater depth. For instance, does the location match the reason for the appointment? Or is someone going long distances for each appointment? If they’re going 200 miles to a dialysis center when there's one that's two miles down the road, we might try to follow up to let them know of closer options. By making this process easier and faster, members can book rides even closer to when they need them, and it allows us to manage benefits more effectively and equitably.  

A Future Fueled by Technology and People

One way people can access their transportation benefit is through public transportation; in fact, many state Medicaid agencies encourage people to use it when available and when they don’t have mobility issues, because it’s cheaper and more effective. We are now working on ways to make it as easy as possible to use public transportation, from helping members with the transit schedule to making sure payment is fast and easy. Doing that at scale across the United States is a very challenging project, because each regional transportation system runs a little differently, but it’s a worthwhile initiative.

SafeRide was built on the premise that a smarter platform produces better outcomes for everyone: members who reach their appointments, health plans that control costs and reduce FWA exposure, and a system that works equitably at scale. As that platform continues to evolve, so does our ability to move more members, more reliably, with less administrative friction. Transportation should be the least stressful part of a patient's care journey, and the right technology makes that possible.

‍