Boxer. Triathlete. One app that finally measures what meditation actually does.
Grady O'Neill is a boxer, triathlete, and endurance athlete who spent years treating his nervous system like a performance variable. HRV before training. Recovery protocols. Everything measured, nothing left to intuition. When he started meditating, it was for the same reason: performance, not peace. Sleep quality. Stress regulation. Recovery speed.
The problem was that every meditation product he tried worked on the assumption that you should just trust it. No measurement. No before and after. No mechanism explained. The interface was soothing. The outcome was unclear. For someone who logs every training variable, this was an unacceptable deal.
In 2022, Grady assembled a small group of athletes, engineers, and clinicians who had been reading the same papers. The mechanism was documented. The technology to deliver it was available. No product had combined them.
In 2023, he went deeper - studying with practitioners who had been training their nervous systems for decades and communities that synchronize group nervous systems through rhythm.
He came back with a working theory of what the product needed to be: a tool that treats meditation the way an athlete treats training. Specific inputs, measurable outputs, documented mechanisms, honest evidence grading.
Mortis started as an argument. Binaural beat entrainment, published by Oster in Scientific American in 1973 and replicated in peer-reviewed journals ever since. MIT's 40 Hz gamma work in Nature, now in Phase III human trials. The Schumann resonance literature. The Stanford cyclic-sighing HRV study. None of it was fringe. All of it was documented.
None of it had been built into a consumer product. Calm gave you a soothing voice. Headspace gave you a cartoon. Insight Timer was a library with no science layer. Waking Up was philosophy. None of them could answer the only question a trained nervous system actually cares about: did it work, and how much did it move the measurement?
The measurable effects of frequency entrainment and long-term meditation on autonomic tone are reproducible and documented. Benson 1975 on the relaxation response. Kaliman 2014 on genomic changes after an 8-hour meditation day. Davidson and Lutz on gamma coherence in Tibetan monks. Iaccarino 2016 on 40 Hz gamma and amyloid clearance. Brook 2013 on TM and blood pressure. Balban 2023 on cyclic sighing. Our job has been to turn forty years of published findings into a tool people can actually use.
Before building anything we spent two years in the field with the practitioners who already had the answers: long-term monks at a Theravada monastery on the Thai-Myanmar border, a tribal community in the Papua New Guinea highlands that synchronizes nervous systems through group rhythm, elite endurance athletes who treat HRV the way their grandparents treated resting pulse, and a short list of neuroscientists who patiently explained where the real science ends and the pseudo-science begins.
What we kept hearing was the same thing in different languages. The athletes said "just track HRV." The neuroscientists said "the frequency-following response is reproducible and documented." Everyone was pointing at the same underlying phenomenon: the nervous system can be moved, on purpose, with the right inputs, and you can measure that it moved.
A voice intake, because the fastest way to read the state of a nervous system is the way the person speaks for thirty seconds. A frequency library, 132 entries deep, because meditation without a specific frequency is a category with no units. Real binaural delivery through the browser and the app, because you cannot entrain a brainwave with a descriptive paragraph. HRV measurement before and after every session through Apple HealthKit, because if it did not move the number it did not happen.
Around that core: a research layer (every recommendation cites the study it rests on), a community layer (the crowdsourced field reports that let us learn faster than any single lab), a breath and posture pairing system, and an honest evidence-tier rating for every node in the library. The ones that are well-supported get flagged well-supported. The ones that rest on tradition get flagged tradition. We never pretend the second is the first.
The name is a direct refusal. The wellness industry has a magic problem. Everything is miraculous, transformative, life-changing. Nothing is explained. No mechanism offered. No measurement possible. You are supposed to just trust that the crystal vibration is doing something.
We are not interested in that. We are interested in what actually happens to a nervous system when it sits quietly with a specific acoustic frequency in its ears for twenty minutes. The answer is: something measurable. Not magic. Physics.
Our library is documented in peer-reviewed literature. Our HRV tracking uses validated methodology. Our recommendations come with citations. We are not claiming to cure anything. We are claiming to deliver a nervous-system intervention you can track, optimize, and use to perform better. If that sounds more like a health tool than a wellness brand, good. That is exactly what it is.
The catalogue began as a research project, not a product. The starting question: which acoustic frequencies have measurable human studies, not just theoretical mechanisms? Every frequency in the library was evaluated against primary literature. Not secondary sources, not meta-analysis summaries, not wellness blogs. Actual papers, human subjects, measured outcomes.
The grading rubric became the evidence tier system: Tier 1 (RCTs), Tier 2 (controlled studies), Tier 3 (observational), Tier 4 (preliminary or animal). Any frequency with nothing above Tier 4 was excluded. Every frequency that made the cut has a primary citation tied to it in the matrix. The total time to build the initial catalogue: eleven months. The number of studies read to get there: 800+.
The 2023 research phase was not planned as product research. It started as a personal question: what does the use of sound and rhythm for consciousness modulation actually look like in communities that have practiced it for generations, before Western wellness extracted and repackaged it?
Studying practitioners across traditions produced one consistent observation: the people who showed the most measurable physiological change across sessions were not the ones with the most experience. They were the ones who practiced at the same time every day, for the same duration, with the fewest variables changed between sessions. Consistency produced the adaptation, not effort.
Communities across the world using acoustic ritual for intentional state change, with no contact with Western meditation traditions, had developed similar timing protocols independently. Fifteen to twenty minutes. Fixed time of day. Rhythmic structure that fell in the theta/alpha boundary range (4-12Hz) when measured by a portable EEG.
"The convergence was not culturally explained. Every community had developed similar timing and frequency structures independently. That is a biological signal, not a tradition."
Grady O'Neill, founder
The product implication: if consistent timing and specific frequency ranges produce measurable convergent effects across unrelated cultures, the mechanism is physiological and it should be measurable in individual users with a consumer HRV device. That is the premise Mortis is built on.
Every product makes tradeoffs that never appear in a feature list. Here are the five that defined Mortis - and the reasoning behind each.
If the product could not show you a before-and-after number, it was just another audio app. HRV via Apple HealthKit was the first API call in the codebase. Every other feature was conditional on this working. If you cannot measure whether a session worked, you cannot build a product that learns from your sessions. The matrix is only possible because of HRV.
A 5-item mood questionnaire forces you to translate your state into categories that were designed for someone else's categories. A voice intake captures how you actually describe your state, in your words, in real time. Claude reads cadence, word choice, and hesitation patterns the same way a skilled clinician would. We lose a small amount of consistency. We gain an enormous amount of signal.
Most wellness apps hide their evidence quality. We publicly label every frequency entry with a tier: RCT, controlled, observational, preliminary. Some of our content is Tier 4. We say so. The competitive disadvantage of being honest about weak evidence is smaller than the long-term cost of users who feel misled. Skeptical users are the most valuable users. We built for them.
We could rank users by RMSSD improvement. We don't. HRV is a biological variable with enormous individual variance - ranking by it rewards genetics as much as practice. Session count rewards the one variable that is fully within your control: showing up. The best-performing practitioners in the beta cohort are the ones who showed up consistently, and the data confirms that. Consistency is the mechanism, not talent.
We could have launched quietly. Instead we scheduled the world's largest single-session binaural HRV collection event and committed to publishing the data regardless of result. If the collective session shows no HRV improvement, we publish that. Zero result, published. This is not a marketing stunt. It is a demonstration of the product's core claim: if we can't measure whether it worked, we have no business selling it. The launch is the argument, not the announcement.
Every one of the five decisions above is a deliberate constraint. HRV tracking means the product can show you that a session did not work. Evidence grading means we have to label our strongest-sounding claims as speculative. Publishing null results means April 18 could become a public proof that this does not work. These are not features. They are commitments that create downside risk for the product and the company.
We made them because the alternative is worse: a wellness product that cannot be questioned, that sells certainty it does not have, that grows until a journalist or regulator forces an accountability moment. The meditation industry has built that product, in many iterations, and it consistently degrades into self-help theatre. Mortis is designed from the product architecture up to not be that.
If enough users measure no HRV improvement, the product's core claim fails in public. We built the measurement in anyway, because a product that cannot survive its own scrutiny should not survive.
The users who stay past 30 days are the ones whose HRV moved. They have personal evidence. That is a different kind of retention than habit or sunk cost. It is the only kind worth building toward.
Calm and Headspace have never published user HRV data. They cannot. Mortis can, and does. That distinction compounds over time into a credibility gap no content-only product can close.
Good. We built this for you. Skepticism is the right default position on a wellness claim. Here is everything we are actually claiming, and the evidence it rests on.
We claim: (1) binaural beat entrainment produces measurable changes in EEG power spectra - documented by Lane et al. 1998. (2) 40Hz gamma entrainment reduces amyloid pathology in preclinical models - documented by Iaccarino 2016 in Nature, now in Phase III trials. (3) HRV is a validated measure of autonomic tone - 40 years of cardiovascular literature. (4) cyclic sighing is the most effective real-time stress-reduction breathing technique tested - Balban et al. 2023, RCT. We claim nothing beyond what those citations support.
Because the founder did not trust the wellness industry's claims, every frequency in the library is graded against primary literature before it ships. No tier inflation. No marketing copy disguised as science.
If you can't measure it, you can't claim it. The decision to require Apple Watch HRV measurement before unlocking the full protocol was a direct product of this skepticism. It forces honesty about whether a session worked.
Tier 3 and Tier 4 protocols are in the library and are labeled as such. We did not remove the lower-evidence content; we told the truth about where it stands. That distinction is the product.
The global meditation and mindfulness market is $9B and growing. Every incumbent in the space sells the feeling of meditation. None of them sells the measurement. Mortis is the only product in the category that treats a session as a data point and tells you, with biometric evidence, whether it worked.
The total addressable market is anyone who already meditates, exercises seriously, tracks recovery, or has been told by a doctor to manage stress. That is not a niche. It is anyone who cares about performance.
SAFE terms, traction data, and the full cap table are available under NDA. If you are a serious investor in health tools, performance technology, or the nervous system space, contact us directly.
Contact for investment deckThe incumbents in meditation are content businesses. Calm and Headspace are libraries. Their defensibility is catalog breadth and brand familiarity. Neither has a biometric feedback loop, not because the technology does not exist, but because adding it would expose a measurement problem they cannot solve: most of their content does not produce a consistent physiological response. HRV tracking would show that. They are structurally unable to measure their own effectiveness.
The fitness apps that already track HRV, Whoop, Garmin, Oura, are hardware companies. They own the sensor. They do not own the content layer and have no reason to build a 132-frequency therapeutic audio library from primary literature. Their business model is the device, not the practice.
Content catalog, no sensor integration, no efficacy measurement. Adding HRV would prove most sessions are inert. Not in their interest to know.
Own the sensor, do not own the practice. Recovery scores are post-hoc. No therapeutic frequency library, no voice intake, no session prescription engine.
Owns both layers: the content (132 frequencies graded by evidence tier) and the measurement loop (HRV before and after every session, read via HealthKit).
Crossing from content-only to measurement-first requires rebuilding the product from scratch: a new intake model, a new evidence framework, biometric integration, and a willingness to surface results that might be negative. That is not a feature a content company adds. It is a different product category.
The second moat is the data. Every session run through Mortis generates a pre-HRV, post-HRV, frequency, breathwork, and time-of-day data point tied to a specific protocol. At scale, that corpus tells you what actually works for what user profile under what conditions. No incumbent is collecting it. No hardware company is correlating it to content. We are building the only dataset that could answer that question.
A beta cohort is only worth running if the findings change the thing you are building. These are the six pieces of beta data that materially shifted the product between September 2025 and launch, with the change each one produced.
The original voice intake asked a structured prompt: "how are you feeling right now?" Users gave 3 to 5 word answers. The signal was thin. When we rewrote the prompt to "tell me what is going on for you, in whatever way you want, for up to 30 seconds," users gave 15 to 25 seconds of content on average, and Claude's recommendation accuracy against HRV outcomes improved by roughly 20 percent.
Change shipped: October 2025. Still in the production prompt.
We originally surfaced the live HRV number in a corner of the session screen. Users started "chasing the number" mid-session, which produced a subtle sympathetic activation and flattened the delta. We moved the HRV display to the post-session screen only. The average delta across the cohort increased by roughly 3 ms RMSSD in the two weeks after the change.
Change shipped: November 2025. Live numbers are now post-session only.
In the first four weeks, beta users ran 40Hz gamma twice as often as delta, even though the cohort had a clear parasympathetic deficit and delta was producing the larger HRV deltas per session. The pattern was driven by stated goal bias: users asked for "focus" and "productivity" in the intake more than "recovery" even when their HRV suggested they needed recovery. We rewrote the recommendation engine to weight HRV data above stated goal when the two conflicted.
Change shipped: December 2025. The engine now overrides stated goal in ~18% of sessions.
When we ranked the beta leaderboard by average HRV delta, users started running shorter sessions back-to-back to inflate their session count. When we ranked by total HRV improvement, users started gaming the baseline. We settled on ranking by "number of sessions completed" which rewards the single variable entirely within user control: showing up.
Change shipped: January 2026. Leaderboard is session-count only.
We originally sent push notifications when users missed a scheduled session. Churn among the reminded users was 2.3 times the churn among users who opted out of reminders. The product was creating shame, not habit. We killed scheduled reminder notifications entirely and replaced them with a morning HRV summary that is only sent when requested.
Change shipped: February 2026. No push notifications for missed sessions.
Beta users could see the evidence tier on each frequency in the library, but not on the frequency they were about to run. They forgot which tier they were in during the session. We added a small tier indicator on the session start screen. Users started self-selecting toward Tier 1 and 2 protocols more often, and the product's average evidence quality per session improved measurably.
Change shipped: March 2026. Tier shows on every session start screen.
Six changes in six months. Each one came from a specific data signal we could not have predicted from first principles. This is what a beta cohort is actually for: finding the assumptions the team did not know they were making, and correcting them before launch. The cohort does not get credit in the App Store listing. They should. The product exists in its current form because 86 people spent 6 months telling us, through their data, what we had gotten wrong.
A product roadmap is only meaningful if it includes what you will refuse to build. The hardest discipline in consumer health is resisting feature creep toward whatever the market will buy next. These are the five directions Mortis will not go, and why.
The product will not add posts, comments, likes, follows, or public profiles. The community data layer is one-way: you see your position in the distribution, you do not interact with other users as people. Why: social features corrupt the measurement. Meditation performed for an audience looks nothing like meditation performed to move an HRV number.
If this is what you want, the wellness category already has excellent options. This is not one of them.
The product will not ship guided meditations voiced by famous actors, athletes, or wellness personalities. The session format is the frequency carrier plus optional breathwork, nothing more. Why: the moment a celebrity's voice enters the session, the measurable effect becomes impossible to separate from parasocial response and expectation. The data stops being interpretable.
The comparison set here is clear. Calm books celebrities. We do not. The choice is structural.
Claude is used for voice intake analysis and recommendation. Claude is not, and will not become, a simulated therapeutic relationship. The product will not add a "your personal AI coach" character who checks in on your emotional state. Why: the parasocial tier of AI companionship is a different product category with different ethical obligations, and we do not want to be in it.
The intake is a tool. It is not a relationship.
Mortis will not seek FDA clearance, will not market to clinicians as a treatment, and will not make disease-specific claims. The product is a wellness tool that measures HRV response to a controlled acoustic intervention. It is not a therapy. We will not soften that distinction to reach regulated markets.
The honest framing is also the legally cleanest one.
Your HRV data will not be sold to advertisers, insurance companies, or aggregators. It will not be used to train third-party models. The only external use of user data is the aggregate community analysis published in Rock Bird, which is anonymized at the session level and never includes identifying information. This is not a marketing statement. It is a hard boundary in the data architecture.
If the company is acquired, the acquirer inherits this commitment or the deal does not happen.
Each of the five directions above is a category where a consumer health product could grow revenue. Mortis will not grow into any of them. The product's value depends on being the narrow, specific thing it is: a measurement loop around a frequency library. Every feature that moves away from that makes the measurement harder to trust. The constraint is the moat.
Mortis is launching globally on April 18, 2026, with a synchronized global meditation event from Cooper Classic, 137 Perry St, West Village, New York, NY. The hook: every participant wearing a heart rate monitor, sharing HRV data in real time, proving a collective wellness claim with data rather than testimonial.
Come to the April 18 launch. Meet the people who want to treat their nervous system like an athlete treats their training.