WILMINGTON, NORTH CAROLINA, August 17 (Hypepotamus) – Is your voice the key to unlocking your health?
Doctor-turned-entrepreneur Morris Nguyen thinks so. A patients’ voice is one of the health signals that is core to his startup Predicate. The platform integrates into healthcare systems to turn everyday health signals, including voice and vital signs, into early insights that help clinicians stay ahead of patient deterioration.
Predicate is starting out with a focus on sepsis, a life-threatening medical condition that impacts 1.7 million Americans a year. For Nguyen, who is building Predicate from his home base in Wilmington, North Carolina, building at the intersection of medicine and artificial intelligence is all about giving physicians more clinical value without creating more work.
Hypepotamus caught up with Nguyen to learn how Predicate works, why voice could become an important health signal, and what comes next as the company moves from R&D into validation and commercialization.
Read through our Q&A with Nguyen here:
Question: How does Predicate fit into a healthcare workflow?
The simplest way to think about Predicate is that we are building an intelligence layer for the healthcare technologies people already use.
Our core technology, OpenDx™, analyzes multiple health signals — including vital signs, voice, symptoms and changes over time — to detect patterns of clinical deterioration earlier. That could mean OpenDx operating inside a wearable, pulse oximeter, remote-monitoring platform, medical device or hospital infrastructure, continuously interpreting signals those systems are already collecting.
The important distinction is that we are not trying to give physicians another application to open or another dashboard to monitor. OpenDx is designed to create clinical value without creating additional clicks. We want the intelligence embedded into the devices, systems, and workflows that already determine how care is delivered.
We think this intelligence belongs closer to where the signals are generated. Our model is closer to a fabless semiconductor company than a traditional health-software company: we develop the IP, models and architectures, and partners implement them in the devices they already build.
Question: Who is making the decision within an organization to start using Predicate?
The strategic shift for us is that the customer is increasingly the organization that builds the device, platform or healthcare infrastructure — not the individual physician who uses it.
For a wearable or medical-device company, that decision may involve product, engineering, clinical and corporate strategy leadership. For a healthcare technology platform, it might be the CTO, Chief Medical Officer or product organization. Within a health system, the conversation can include clinical informatics, digital health, innovation and clinical operations.
The common question is essentially the same: How do we make the technology we already have more intelligent?
Predicate can provide that intelligence without requiring a partner to build the underlying signal-processing, clinical reasoning, and AI models themselves.
Our long-term vision is that a partner should be able to embed OpenDx into something they already make — whether that is a device, platform, or piece of healthcare infrastructure — and provide evidence-based context about physiologic change, not just collect data.
Question: For a non-medical reader, what makes voice signals so interesting to follow?
Voice is fascinating because it is an incredibly information-rich biological signal that we generate naturally every day.
Speaking requires coordination across breathing, the vocal system, muscle activity, cognition, and language. When something changes physiologically, it can subtly affect characteristics of speech and voice — breathing between phrases, cadence, vocal quality, energy, timing and other features that may not even be obvious to the human ear.
And unlike many medical measurements, collecting voice can be almost frictionless. You don’t necessarily need a blood draw, laboratory, or specialized piece of equipment. In many situations, you simply need someone to speak.
But what really interests us is not voice in isolation. One of the foundational ideas behind OpenDx is signal fusion. Voice becomes far more informative when it can be interpreted alongside vital signs, symptoms, an individual’s baseline, and how all of those signals are changing over time.
I often describe it this way: voice is another sensor — except it is a sensor the human body produces naturally.
Finley Raises $1.85M for Middle-Market Finance
Wilmington FinTech Finley raised its pre-seed round to bring AI-powered financial intelligence and CFO-level insights to middle-market companies.
Question: Have you raised outside capital?
Yes. Predicate has raised under a million in outside capital from a small group of private investors who backed us very early.
We’ve been extremely capital-efficient with that funding. It has supported the development of OpenDx, our data science and signal-intelligence work, international research collaborations, the scientific and clinical network surrounding the company, and the intellectual property that recently resulted in U.S. Patent No.12,694,957.
We are now entering a very different stage of the company. The foundational technology has been built, the intellectual property is developing, and our research footprint is expanding internationally.
Additional capital now becomes acceleration capital — allowing us to move faster through validation, hardware architecture and reference design work, strategic partnerships and commercialization.
The Wilmington technology community has changed tremendously over the past five years.
There is a genuinely supportive ecosystem here — experienced entrepreneurs willing to mentor younger companies, a strong network of people who want to see one another succeed, and organizations like the Center for Innovation and Entrepreneurship at UNCW that have helped create an environment where founders can build.
For Predicate, winning the 2025 NCTECH AI, ML & Data Innovation Company of the Year award was particularly meaningful because it demonstrated that a company developing sophisticated AI technology from Wilmington can compete and be recognized on a statewide stage.
I think Wilmington’s next major opportunity is increasing the density of deal flow and growth capital around the companies being created here.
Healthy technology ecosystems eventually become self-reinforcing: strong companies attract capital; capital allows those companies to hire, grow and remain in the community; successful companies then produce experienced founders, executives, employees and investors who support the next generation of companies.
I think Wilmington is much closer to that flywheel than people outside the region realize.
wilmington – Hypepotamus
Your news source for tech in the Southeast. Original, local reporting on startups, investors, and the corporate innovators shaping the region.
Question: Tell us a bit about the team you’ve assembled to build Predicate.
One of the things that makes Predicate unusual for a company at our stage is the depth of expertise we’ve assembled around the science and engineering.
Our team spans signal processing, data science, artificial intelligence, clinical medicine, and translational research.
Our signal-intelligence work is led by an exceptional signal-processing engineer, supported by a data science team developing the mathematical and machine-learning models that allow OpenDx to interpret multiple physiological and contextual signals together.
We’ve also surrounded the company with world-class clinicians and researchers across critical care, medicine, global health, and related fields, including an internationally recognized voice-biomarker researcher helping us advance the science around what can be learned from human voice.
That combination is important because what we’re building requires much more than a good AI model.
You have to understand the signals. You have to understand the mathematics. You have to understand the hardware. And ultimately, you have to understand the clinical problem you’re trying to solve.
Predicate has intentionally been built at the intersection of those disciplines.
Question: What are you focusing on for the rest of 2026?
The rest of 2026 is about turning several years of R&D into evidence and partnerships.
We currently have six validation programs underway or initiating across five countries, and one of our major priorities is to initiate all of our planned validation programs and complete several of them this year.
Those studies allow us to evaluate OpenDx across different patient populations, clinical environments, and healthcare systems and continue building the evidence required for broader deployment.
At the same time, we’re advancing the next evolution of OpenDx: moving this intelligence closer to the point where health signals are generated.
That means taking AI models that can detect patterns of clinical deterioration and developing the specialized computing architectures that could ultimately allow that intelligence to operate at the edge — inside medical devices, wearables, and the infrastructure healthcare systems already depend on.
That is where the semiconductor strategy becomes particularly important. Predicate doesn’t need to own a semiconductor fabrication plant. As a fabless company, our opportunity is to develop the intellectual property, AI models and hardware architectures, then work with semiconductor manufacturing and strategic partners to bring those designs into the physical world.
The next phase is proving it at scale and putting OpenDx into the hands of partners that can distribute that intelligence to millions of people.
