The questions serious institutions ask first.

URIEL™ is institutional decision intelligence: a system that helps people hold their judgment steady in the environments where losing it is most expensive.

No. Coaching tools are built for individuals who have opted into personal development. URIEL is built for institutions that carry consequence — hospitals, agencies, commands, and boards — and it is deployed under the institution’s governance, inside its data protection posture, on its terms.

Contemporary AI gives an organisation speed and computation. It does not give it restraint. Under sustained pressure, capable people do not become less intelligent — they become less deliberate. Decisions migrate from considered judgment to reaction. URIEL exists to protect the deliberate part.

Environments with no margin for error. Oncology is the first deployment lane, where clinician load, moral weight, and patient adherence intersect more sharply than almost anywhere else in medicine. The same architecture extends to broader healthcare, law enforcement, command, and the families and caregivers who absorb the second shift of all three.

Most of them are optimised for engagement, output volume, and speed. URIEL is optimised for decision quality, and it is willing to be slower to get there. It is designed to introduce friction at the exact moment friction is valuable — which is the opposite of what a productivity tool is built to do.

Emotional intelligence tells you what state you are in. Ethical intelligence governs what you do next while you are in it. One is instrumentation. The other is a brake and a steering column. Knowing you are depleted has never, on its own, stopped anyone from making a depleted decision.

No. URIEL is model-agnostic by design — the language model is a component, and it is replaceable.

What sits above it is not: eighteen validated frameworks from behavioural and clinical science, integrated into one decision architecture, together with the constraints governing what the system may say and when it is required to stop.

The science is peer-validated and public. The integration is Dr. D. Ivan Young’s — two decades of applied practice with people deciding under real consequence.

The language model supplies the words. URIEL supplies the judgment. Only one of those is available for licence.

Nothing stops them building something adjacent. What is hard to acquire is the position.

Three things have to be true at once. The method has to come from two decades of applying it to people deciding under real consequence — practitioner knowledge, not training data. It has to clear clinical timelines, which capital cannot compress. And it has to be trusted by the governance, security, and procurement committees who are, correctly, most sceptical of the companies best resourced to build it.

There is also a structural problem. A horizontal platform is rewarded for engagement. A system whose value shows up as decisions not made and harm not caused is not what those incentives produce.

And the work does not generalise. Building one is a project. Building a dozen, each of which has to survive interrogation by the people who actually do that job, is not twelve times the same project. A rifle company and an oncology ward do not find the same paragraph credible. Every environment is earned separately, and the ones already earned do not make the next one cheaper.

This matters more than anything else on this page, so we will be plain about it.

URIEL is not a medical device. It does not diagnose, treat, cure, or prevent any condition. It is not a therapist, not a clinician, and not a crisis service. It does not make decisions on anyone’s behalf, and it is not a substitute for clinical judgment, command authority, or professional supervision. Any deployment that would require it to be any of those things is a deployment we decline.

Generative output is bounded rather than open-ended, and the boundaries are set by the underlying behavioural frameworks rather than by the model’s own confidence. Inputs that signal clinical, legal, or safety risk do not receive a generated response at all — they are routed out of the generative path and into a defined escalation, with a human being at the end of it. The system is built to recognise the edge of its competence and stop, which is a design requirement, not a limitation we are apologising for.

The components are established and peer-validated — eighteen recognised frameworks from behavioural and clinical science. We did not invent the science and do not need to. The evolution is in how it is integrated.

Dr. D. Ivan Young’s research on self-empathy as a trainable clinical skill has been accepted for presentation by the American College of Lifestyle Medicine at Lifestyle Medicine 2026, and his book Leading from the Heart is published by Post Hill Press with Simon & Schuster distribution.

We publish no efficacy percentages. Clinical validation is underway. We would rather be trusted later than quoted early.

Under an executed agreement, an institution’s clinical, security, and governance reviewers get the access they need to satisfy themselves — including the reasoning path behind any given output. We do not publish that architecture on a website. Explainability to a customer under diligence and public disclosure of proprietary method are different obligations, and we take the first one seriously precisely because we do not confuse it with the second.

Nowhere it does not have to. URIEL is deployed inside the institution’s existing data protection posture, under its governance, with defined retention. We do not sell data. We do not train shared models on identifiable institutional data. In the United States that means HIPAA-aligned handling; in the United Kingdom, UK GDPR and the Data Protection Act 2018, the Caldicott Principles, and the NHS Data Security and Protection Toolkit.

The underlying science does not change — pressure degrades judgment the same way in a treatment room and an operations centre. What changes is everything above it: the stressors, the compromises, and the limitations native to that environment.

And the differences run below the level of sector. What sits on a soldier in the hours before an assault is not what sits on a sailor six months into a deployment with no horizon that changes. Neither resembles a paramedic at hour nineteen of a tour, or a dispatcher who has just taken a call she will not be told the ending of.

Calibration at that resolution is done per environment, with people who have worked inside it. Not per sector, from the outside.

No. URIEL is deliberately positioned as an enhancement of human agency rather than a substitute for human authority. That is an ethical commitment, and it is also the reason the system fits inside regulatory frameworks that are tightening around autonomous decision systems.

Compliance was designed in at the start rather than retrofitted. The architecture anticipates the EU AI Act and comparable US federal oversight, and the non-diagnostic, human-in-the-loop design is the reason it does so from a position of strength.

The question assumes URIEL issues something to act on. It does not. It is non-diagnostic, recommends no course of treatment, and gives no instruction — authority remains with the clinician, the commander, or the supervisor, and the decision record stays where it already lives.

Liability is allocated in the deployment agreement, negotiated with the institution’s counsel alongside the rest of the terms. It is not something we assert on a website.

Not by default, and most early deployments will not. URIEL does not require access to the clinical record to do its work. The lightest integration is usually the right one — every system that touches the record adds months to security review and considerably more to change management.

Where an institution wants integration, it is scoped through the same review as any other system, on existing standards and existing identity infrastructure.

SOC 2 certification is in progress. We will say that plainly rather than imply more, and reviewers are welcome to ask where in the process we are.

In the interim, deployments run inside the institution’s own security perimeter and are scoped against its existing controls, which is the posture most information security teams prefer at this stage regardless of what a vendor holds.

URIEL is developed by Young Ethical Intelligence, Inc. It was founded by Dr. D. Ivan Young, a behavioural scientist and master-level practitioner whose applied research examines the neurobiology of judgment under pressure. Our Chief Strategic Medical Advisor is Ambaw Bellete, formerly President and Chief Operating Officer of CG Oncology, Inc. (NASDAQ: CGON).

URIEL is in active development, with oncology as the first deployment lane, and in active conversation with private sector and institutional parties about scoped early deployments.

We are not claiming customers we do not have, and we are not naming the ones we are talking to. Institutions that entrust us with the judgment of their people are entitled to the same discretion we extend to individuals. What we discuss, we discuss under NDA, with clinical leadership, security, and procurement in the room.

Through the institutional contact route. Deployments are scoped with the institution, under its governance, and inside its data protection posture — so the first conversation is usually with clinical leadership, security, and procurement in the room together, rather than a product demo.