Consider reading these articles
.webp)
.jpg)
No items found.
HLTH 2025 Big Takeaways
.png)
No items found.
D2C Care Navigation: Healthcare's Missing Ally Emerges
No items found.
Dan Walmsley’s first major system out of college was an air traffic control platform deployed to Incheon Runway 3, the busiest runway in the world. His most recent system powers an AI companion he built for his mother in a rural nursing home in Queensland, one that remembers her history, recalls prior conversations, and helps connect her with family across time zones.
Scott Hoch was a physicist at Yale, where he built experiments to measure phenomena at the limits of what can be observed, working with quantum systems, photons, and structures just atoms thick. He left academia to build in the real world, and since then has developed production systems across healthcare, data infrastructure, and AI that are still running years later.
That combination says a lot about how they think and build as founders, and it’s what made us so compelled to back Dan (Co-Founder and CTO) and Scott (Co-Founder and CEO) and lead Gravity Rail’s seed round, an AI platform for scaling patient engagement while giving care teams full control over their workflows.
The founders
Dan credits much of how he builds to his parents. His mother, Joan, spent decades helping children with disabilities participate in mainstream education at a time when many were excluded from it entirely. His father, John, was a product design engineer who built everything end to end and, after seeing failures in hang gliding safety, helped rewrite standards for the sport. That orientation toward building systems that actually work for the people using them shows up clearly in Dan’s own career.
As the founding CTO of NationBuilder, he built the infrastructure behind what became the operating system for political organizing worldwide, now serving more than 110,000 customers across 110 countries. The philosophy was straightforward: give the operator control. Let the people closest to the work shape how it gets done.
He spent the next decade at Automattic, rising to Principal Architect for AI. During that time, he repeatedly took on projects that would typically require entire teams. Working largely on his own, he built Jetpack Boost, a performance product for WordPress sites that became the number one growth driver for the business that year. He also re-architected the company’s flagship WordPress plugin, used by millions of sites, from a monolith into modular packages, a project that would ordinarily take a team months. In a single afternoon, he shipped a partner login portal prototype that now serves more than 40,000 agencies.
In our diligence, his peers struggled to find a clear comparison point. Former colleagues used phrases like “beautiful mind” and “mad scientist.” One placed him in the top few percent of technologists he had ever worked with for raw productivity. Another said that in a company full of elite distributed systems engineers, Dan operated on a different plane.
But what came through just as clearly was something harder to quantify. People don’t just work with Dan, they follow him. NationBuilder’s co-founder described his magnetism and charisma as a recruiting advantage, and it shows in the team he has built at Gravity Rail. Several former colleagues chose to join him again, drawn by how he works and what he expects from those around him.
Scott started his career measuring phenomena at the limits of what can be observed. At Yale, he worked at the boundary of quantum systems and classical telecommunications, building experiments to measure noise in laser signals caused by the random arrival of individual photons. He built structures just ten atoms thick and measured motion at scales that are difficult to observe, let alone control.
At NationBuilder, he helped build a national voter file, stitching together fragmented data across thousands of sources and solving identity at scale. At Deep6 AI (later acquired by Tempus), he built an NLP negation model that outperformed the company's third-party vendor for matching patients to clinical trials across complex healthcare datasets. At Revenue.io, he architected data systems powering real-time conversation intelligence across thousands of concurrent calls, where references described him as a "force multiplier." Through his own firm, Black Box Engineering, he developed a kidney failure detection algorithm for a value-based care organization that has been in continuous use since 2020 — the same organization that later became one of Gravity Rail's first customers.
Across each of these, the common thread is execution. Colleagues described someone who could take an ambiguous problem, understand it quickly, and turn it into a working system almost immediately. One described explaining a client's need in the evening and waking up the next morning to a functional prototype.
Dan and Scott first met more than a decade ago at NationBuilder, where Dan recruited Scott out of his PhD. Their careers diverged after that, but they stayed close. When Dan decided to start Gravity Rail, Scott was the first person he called.
The team today is six people, two founders and four engineers, all of them technical. Scott, by every measure a strong technical operator, is also the least technical person on the team — he and Dan joke that Scott isn't allowed to touch the code. The engineering density is unusual, and it's what allows them to define complex systems and build them quickly with a standard that holds as they scale.
Why this problem, and why Gravity Rail
Care management is one of the highest-leverage functions in healthcare, and one of the most broken. Patient engagement rates are dismal. Care managers spend their days on administrative work instead of clinical judgment, and patients fall out of programs because no one followed up in time. Human capacity simply doesn’t scale to meet the need.
That NationBuilder instinct — give the operator control — is what led Dan and Scott to see something healthcare insiders had largely accepted: the SaaS model for care management gives you someone else’s idea of what your workflow should be. It scales the “expert’s” opinion, not the care manager.
Over time, that creates a layered problem. Care teams are forced into rigid systems that don’t reflect how they actually work, while managing a patchwork of vendors, data systems, and integrations. The result is more complexity, not less.
Dan and Scott had already shown what it looks like to build infrastructure that puts the operator in control at global scale, and Gravity Rail extends that same philosophy to patient engagement.
Operations teams can translate their own SOPs into live, multi-channel workflows using natural language, without code or implementation vendors. The platform handles high-volume interactions across enrollment, outreach, medication adherence, and routine communication, and escalates to the care team when judgment is required. The organization owns the workflow and can evolve it over time.
Underneath that experience is a real-time orchestration engine built on an architecture Dan designed, allowing the system to handle tens of thousands of patient interactions without breaking, while running voice interactions at a fraction of the typical cost.
What was already working
By the time we met Scott, the platform had been in production for a few months. In a chronic kidney disease program at a large academic medical center, patients remained engaged beyond twelve weeks, with engagement increasing over time. This wasn’t because the system nagged patients about compliance, but because it talked to them about their lives, learned what mattered to them, and built the kind of genuine trust that no rigid protocol can manufacture. The value-based care organization running that program estimated a 5x efficiency gain for care managers, with a roadmap to expand across their full 25,000-patient population.
Leapcure, a clinical trial recruitment company, stress-tested the platform at scale. They landed a flu vaccine clinical trial requiring 15,000 recruits, 10x their normal scale. They began with a cohort of 3,000 patients, then expanded to the full trial as results came in. The sponsor ultimately doubled the allocation to 30,000.
Along the way, Leapcure improved patient conversion rates by 20 percent while operating with a leaner coordinator team. Engagement no longer depended on business hours or manual follow-up. By the end of the trial, Leapcure’s team was self-sufficient on the platform, building workflows for their next trial without any outside support.
What we’re betting on
Both founders came to this problem personally. Dan was coordinating care for his mother through a period of deteriorating health that left her in a nursing home in Brisbane. At the same time, Scott was navigating his own father’s health crisis. In many ways, this was an idea 30 years in the making, shaped by both founders’ experiences navigating care for their own parents.
Redesign Health invested $2.75 million to lead Gravity Rail’s seed round because of a specific convergence: a founding team with a proven track record building platform companies at global scale, now pointed at a problem they understand personally. The product architecture addresses the real barrier to adoption in healthcare AI, control, and early production-grade traction from a small team proves the architecture works.
Healthcare operators are becoming builders. We saw it in the CKD care managers who evolved their own workflows, and in the Leapcure team that launched a new trial without picking up the phone. At Redesign, we believe the platforms enabling that shift will define the next generation of care delivery. Gravity Rail is the first one we’ve seen that gets it right, and we’re proud to be a part of it.
DATA INSIGHTS
Return to insights