Why Now
The best medtech is positioned for something new
The most successful medtech companies have hardware platforms that work. They solve unmet medical needs. They’ve earned trust the slow way — through quality systems, clinical evidence, and years of devices performing where it counts.
And now they can move into the space of rapid software innovation.
That sentence is easy to write and hard to earn. Rapid software iteration on a regulated, safety-critical platform is one of the most demanding engineering problems in any industry — which is why the companies that can do it are so hard to catch. Only the best medtech can really capitalize on software throughput. The platform, the quality discipline, and the install base are the prerequisites, and they don’t come from a tool. They come from having done the hard part already.
Where the compounding happens
Throughput as the frontier
Hardware platforms set the ceiling; software throughput determines how fast you reach it. Every improvement a team can validate and ship — a better algorithm, a refined interface, a closed safety gap — compounds across an installed base that hardware-only competitors have to rebuild device by device.
The constraint is rarely invention. Most organizations can design a change faster than they can prove it’s safe to ship, because each change must be validated against an increasingly complex integrated system. The companies that compress that proof cycle — without giving an inch on quality — set the pace for everyone else.
That compression is what a new class of software-understanding technology aims at: answering “what behavioral change did this build introduce, and with what confidence?” so testing effort can focus on what actually changed. Whether a given tool delivers on that for a given manufacturer is an evidence question. That’s the question MTL exists to answer.
Patch velocity makes the case sharper
The same capability, under a harder clock
There is a second reason throughput matters, and it arrives on someone else’s schedule. Security researchers and federal program leaders describe a window from public vulnerability disclosure to active exploitation that has collapsed from years to days. In that environment, the ability to evaluate, validate, and ship a fix quickly is part of what it means to be responsible for devices in the field.
The encouraging part: this is the same capability, not a separate one. The throughput that ships features fast is the throughput that ships fixes fast. Manufacturers investing in validated software speed are strengthening both at once — and they deserve evidence about which investments actually move the clock.
Serious R&D money is already at work on this
The federal R&D context
ARPA-H has invested $90M+ in software-understanding technologies — tools proven in domains like defense and large-scale software, now being directed toward health applications. The performers building these technologies are funded to demonstrate real value with real manufacturers, not to run sales demos.
Medtech Throughput Labs is preparing a submission to ARPA-H’s FY26 SBIR solicitation (Lineage topic) to run the independent evaluations that connect those performers to manufacturer reality. MTL is paid by the program, not by the performers — the evaluations exist to produce a fair answer, whichever way it points.
The first evaluation cohort is forming now; evaluations proceed under federal SBIR funding upon award.
Why this is worth 30 minutes now
The close
The cohort is small — three manufacturers, evaluated deeply — and it is being assembled from the companies with the most to gain: those already investing real effort in throughput, where an evaluation turbocharges work their teams are already doing rather than adding something new.
If that describes your organization, a 20–30 minute briefing is enough to see whether the fit is real. If it is, your team gains a benchmark and an ROI model it keeps regardless. If it isn’t, you’ll know quickly, and the conversation will have cost you half an hour with people who take your constraints seriously.
Either way, the work continues toward the same end: that great tech gets all the way to the people who need it most.
See whether the fit is real
A 20–30 minute briefing with Shannon — your questions drive it. If your organization is already investing in throughput, half an hour is enough to know.