Smartwatches have become remarkably good at recording the heart. Interpreting what they record is another matter. Anyone who has taken an on-wrist ECG and received an “inconclusive” result knows the gap between capturing a signal and making sense of it. On November 5, 2026, a Belfast company called PulseAI announced that it had received U.S. Food and Drug Administration (FDA) 510(k) clearance for an AI-powered arrhythmia analysis model built specifically for wearables, a step aimed squarely at closing that gap.
According to PulseAI’s announcement, the model improved detection sensitivity for atrial fibrillation (AFib) in a published study and cut inconclusive readings by 92% compared with the Apple Watch’s built-in algorithm. For a market flooded with consumer heart data, that is a notable claim, and the clearance gives it a regulatory footing.
What PulseAI actually announced
PulseAI describes itself as a medical technology company developing artificial intelligence for cardiac diagnostics. Its press release states that the FDA has cleared its “AI-powered arrhythmia analysis model for wearables.” A 510(k) clearance is the FDA pathway through which a company shows its device is substantially equivalent to a product already legally on the market; it allows the product to be sold in the U.S. for its cleared purpose.
The product is not a watch. It is software that sits behind other companies’ devices. According to PulseAI, the technology is designed to integrate with compatible ECG devices and software platforms: recordings are sent to the cloud for automated analysis, and structured results come back to the connected app for clinical review. In other words, PulseAI is positioning itself as an interpretation layer that device makers and care providers can plug into, rather than as a consumer brand.
The company was founded in 2020 and is based in Belfast, Northern Ireland. Its release says its deep learning algorithms were trained on millions of ECGs and are intended to perform across smartwatches, patches, hospital monitors and defibrillators.
The problem it targets: too much data, too little clarity
PulseAI frames the opportunity around scale. Citing IDC figures, the company notes that 164 million smartwatches were sold worldwide in 2025, and most of them offer some form of heart rhythm monitoring. That pushes ECG recording far outside hospitals and clinics, but it also creates a new workload for the clinicians who are asked to look at those recordings.
The release singles out inconclusive results as the biggest pain point. An unclear reading can trigger manual review, a repeat recording and a follow-up with the patient. Each of those steps costs staff time. PulseAI’s argument is simple: if software can resolve more recordings confidently, clinical services can handle more patient-generated ECGs without hiring proportionally more people to read them.
The clinical stakes explain why AFib is the focus. As the company’s announcement notes, AFib is the most common sustained heart rhythm disorder, affecting tens of millions of people worldwide. It can occur without symptoms, and when untreated it is associated with roughly a fivefold higher risk of ischaemic stroke, depending on individual risk factors. Identifying it is what allows clinicians to assess stroke risk and decide on next steps.
The evidence behind the headline numbers
The key data point comes from a prospective study of 296 participants published in the journal Heart Rhythm. According to PulseAI, its algorithm improved AFib detection sensitivity in that study while producing a 92% reduction in inconclusive interpretations compared with the Apple Watch’s native algorithm. The company’s subheadline goes further, saying the published work shows “up to 10x performance improvement when used with leading wearables.”
Two caveats are worth keeping in mind. First, the numbers come from the company’s own summary of the study; readers who want to weigh the methodology should look at the Heart Rhythm paper itself. Second, the comparison is against one manufacturer’s on-device algorithm. Wearable makers update their software regularly, so the size of any advantage could shift over time.
Still, a 92% drop in unreadable results speaks directly to the workflow problem PulseAI describes. A tool that sharply reduces the “we can’t tell” pile is valuable even before you consider sensitivity gains.
Funding and the university connection
PulseAI’s release says Innovation Ulster Limited, the investment arm of Ulster University, led its recent investment round alongside the Mortara Group. Ulster University’s Vice Chancellor, Professor Paul Bartholomew, described the company’s path “from University spin-in to securing FDA clearance” as a significant achievement for the company, the university and Northern Ireland.
Founder and CEO Dr Alan Kennedy summarized the company’s goal in the announcement: “Wearables have made ECG recording widely accessible. Our ambition is to match expanded access with improved accuracy to help deliver a truly scalable solution to providers.” He added that, with clearance in hand, the focus now is “bringing this capability into the devices and platforms in use by millions every day.”
What it means
For readers following health technology, this clearance points to a few broader shifts:
- Software is becoming the product. PulseAI does not need to build hardware. Its business depends on licensing interpretation to the companies that already own the wrist, the patch or the monitor.
- Accuracy is the new battleground. Capturing an ECG is now routine. The competitive edge is in reducing ambiguous results and the clinical labor they create.
- Clinician workflow drives adoption. The pitch is aimed at providers and platforms, not consumers. Results are returned “for clinical review,” which keeps a professional in the loop.
- Partnerships are the next milestone. The announcement does not name the specific wearable or platform partners that will ship the model. Watch for those deals; they will determine how many people ever encounter the technology.
This is product and regulatory news, not guidance on personal health decisions. Anyone with questions about their own heart rhythm readings should raise them with a qualified clinician.
Conclusion
Wearables solved the access problem for ECGs; they did not solve the interpretation problem. PulseAI’s FDA clearance is a bet that a dedicated, cloud-based AI layer can turn a noisy stream of wrist recordings into results clinicians can actually use. The published data on inconclusive readings is promising. The real test will be whether device makers and health systems integrate it, and whether the gains hold up at the scale of millions of users.