Define what is permitted.
Translate human intent into explicit limits. Technical capability must not silently expand permission.
The control system behind KeyRD
More capable systems.More human control.
Statera is building the foundations for software that acts within clear human authority—and remains answerable for what it does.
A long-term development effort.
KeyRD is our first product commitment: build deeply enough in dietetics to deserve a durable place in practice.
The governing idea
Human intent & explicit authority
The control system
StateraContext · Permission · Boundaries · Evidence
A permitted action, a useful question,
or a deliberate pause.
or a deliberate pause.
Capability is not permission.
From intelligence to accountable action
A capable assistant still needs a clear boundary.
The model is only part of a system. Our work is the structure around it: what may happen, what evidence is sufficient, and when control must return to a person.
Know what is known.
Distinguish current evidence from stale information and uncertainty. Missing facts are not an invitation to guess.
Choose the least needed.
Act within the boundary, ask a useful question, remain quiet, or return the decision to a person.
Leave a reason to inspect.
Make decisions reviewable. Build toward meaningful challenge, stopping, and correction—not just explanation.
The first commitment
A broad aim.
A profession we intend to serve deeply.
KeyRD
Keep patients connected to care—with less chasing.
A continuity assistant under development for RDN-led nutrition care. The goal is to extend RDN expertise between visits without inventing professional authority.
Explore KeyRDWhy dietetics is our first commitment
Professional expertise. The Nutrition Care Process. Real-world choices outside the appointment. And a clear obligation: build something useful enough for RDNs to keep, not merely a demonstration that teaches Statera what to do elsewhere.
Dietitian-delivered care has a clinical evidence base, while reviews of digital dietary support show both promise and important limitations. Those findings are directly relevant to why KeyRD begins with professional nutrition care. [5] [6] [7]
Why dietetics is where we beginEvidence closest to the first product
Start with nutrition care before zooming out.
The immediate question is not whether AI can help somewhere. It is whether carefully bounded software can improve the work around RDN-led care.
Professional nutrition care has evidence behind it.
Our first responsibility is to preserve the value of dietitian-delivered care rather than substitute a generic agent for it. [5]
Digital support can help, but results are mixed.
Reviews motivate testing between-visit support without assuming that another digital touchpoint improves adherence or outcomes. [6]
Context and implementation matter.
Condition-specific evidence reinforces why the product has to be evaluated inside real RDN workflows rather than inferred from generic AI performance. [7]
The broader research direction
Human control has to be real, not ceremonial.
Expertise can sometimes reach further.
Workplace evidence shows AI assistance can improve productivity in some settings; that does not establish benefit in dietetics. [1]
Approval alone is not a safety system.
A meta-analysis found human–AI combinations improved on humans alone but underperformed the better standalone alternative on average. Workflow design and testing matter. [2]
The objective matters.
A health-algorithm audit showed how the wrong proxy can produce unequal harm even when the system appears to optimize successfully. [3]
The 20-year aim
Greater capability.
Without diminished agency.
Build shared foundations that keep consequential machine actions authorized, justified, contestable, stoppable, and repairable.
The human purpose is practical: more access to useful expertise, more time for meaningful work and relationships, and more—not fewer—choices about how to live.
Read the long-term aimHow the work advances
Commit here before we generalize.
Build deeply enough to deserve a place in practice.
Compare KeyRD with credible simpler alternatives. Measure appropriate action, restraint, patient choice, practice workload, and the conditions RDNs need for continued use.
Learn without abandoning the first commitment.
Evidence from KeyRD may inform later work, but another setting needs its own boundaries and proof. Generalization is additive; it is not a plan to discard the dietetics product.
Make delegated power accountable.
Develop reusable control methods while maintaining responsibility for products already placed into practice.
Who is accountable
A named founder. A broader company thesis.
Statera Systems Inc. was founded by Neil James.
The personal path that led James through nutrition science, clinical research, and dietary-adherence evidence to KeyRD is told where it is most relevant: on KeyRD’s site. Statera is responsible for the broader work: building and testing the control system, owning the operating boundaries, and remaining accountable for the claims attached to its products.
James holds a BS in Nutrition & Food Science from LSU and completed the didactic program in dietetics. He is not an RDN. Practicing RDNs are informing KeyRD’s workflow and boundary evaluation. Advisors will be named publicly only when a defined role and permission to be identified are in place.
Why KeyRD became the first productExplore the work
A long-term argument.
A first product you can inspect.
Start with KeyRD or read the evidence behind the direction.