Semantic Computing Research

The industry is generating entropy. We are building guarantees.

Computing lost its semantic layer. We are restoring it.

8 Patents · Enterprise Validated · Production Deployed
The Paradigm

A new category of computing where meaning is the primitive.

Semantic computing treats domain knowledge as a first-class computational artefact — formally modelled, mereologically coherent, and deterministically enforced. Your problems are reified into structures that are formally guaranteed — not approximations a developer maintains.

Without semantic computing
AI refactors GL Account Code Types
The AI shortens “CL” to “C” — it looks cleaner, and all tests pass. Every downstream report, tax submission, and bank reconciliation now uses the wrong code. Tests verified syntax. Nobody verified meaning.
Result: silent semantic failure
With semantic computing
The specification knows the code’s meaning
The GL code is formally bound to a regulatory codebook. The AI cannot shorten it without violating the invariant. The guarantee is structural — not procedural, not dependent on test coverage, not a convention someone remembers.
Result: violation prevented at specification level
Semantic
Meaning preserved through all transformations. Intent in, correct behaviour out.
Ontological
Domain structure formally modelled. Declared, not implied in code.
Mereological
Part-whole coherence guaranteed. Every part preserves the truth of the whole.
Reification
Abstract problems become concrete, enforceable, verifiable structures.
AI-First
Built for AI to understand, evaluate, and operate — not retrofitted.
Deterministic
What you specify is what runs. Not stochastic. Not probabilistic. Guaranteed.
The Thesis
The specification appreciates.
The application depreciates.

In the AI era, code is generated at zero marginal cost. Code is the commodity. What remains irreducibly valuable is the domain specification — the business knowledge that governs correctness. Every refinement of a constraint, every clarification of an invariant, every enrichment of a policy makes the specification more valuable.

Lines Of Stochastic Source produce entropy. Semantic specifications produce compounding knowledge. One depreciates. The other appreciates. The entire industry is accelerating toward entropy. We are founding the discipline of appreciation.

The Evidence

We defined a paradigm.
Then we built it.

The working system is called AYIOS — a semantic computing platform where one typed specification simultaneously becomes the running application, its enforcement layer, its access control, and its AI guidance. All coherent. All from a single source of truth.

Production Case Study
Enterprise wealth management platform — semantically reified from a legacy codebase into formally guaranteed specifications.
4,163
business flows reified
269
semantic modules
284K
lines of specification
100%
compilation pass rate
8
Patents
96 claims · 3 USPTO (Patent Pending)
5/7
Enterprise Dimensions
Empirically verified · reproducible
97%
Memory Reduction
Production · water infrastructure
0
Panics
237 adversarial schemas submitted
The Working System AYIOS → Explore the semantic computing platform Mendix Intelligence mxto.ai → AI-powered semantic analysis for Mendix applications Latest Research Why You and Your AI Are at a LOSS → The endemic failure mode in AI-assisted code generation
The Position

Three quadrants produce stochastic output.

One produces deterministic guarantees.

← Specification Code →
Formal Methods
Dafny, TLA+, Coq — rigorous proofs, but no running applications.
Semantic Computing
Domain knowledge formally preserved through execution. Deterministic. Appreciating.
SDD + Governance
Kiro, Tessl, Credo AI — spec guides code, but code remains stochastic.
AI Code Generation
$50B+ in valuations. Lines Of Stochastic Source at unprecedented speed.
↑ Appreciating ↓ Depreciating

Every alternative produces stochastic output with no formal relationship to domain knowledge. Only semantic computing treats the specification as the execution itself.

The Institution

Ontology Labs

Founded on the thesis that computing lost its semantic layer — and that restoring it requires a new formal foundation: ontological structure, mereological coherence, and deterministic reification of domain knowledge.

Founded by Hardy Jonck. From defence simulation and control systems to semantic computing — three decades of modelling complex domains from first principles. Eleven ventures. One culmination.

Our research programme: defining the formal foundations of semantic computing through ontological structure, mereological coherence, and deterministic reification — and proving them in production.

8 Patents · 96 Claims 3 USPTO Filed Enterprise Validated Production Deployed
Engage
Engage directly

A conversation about semantic computing for your organisation.

Get in Touch or schedule a call →
Let your AI evaluate

Structured evidence endpoints. No sales pitch required.

AI Evaluation
Explore the platform

AYIOS — the working semantic computing platform.

Visit AYIOS