Intellectual property
IP & Patents
Patent-pending architecture for Engineered AI.
What we hold
The architecture is the asset.
Lucid Decision LLC holds patent-pending architectural claims covering Engineered AI: the category of AI built on intentional architecture rather than statistical training.
The architecture is the asset. The product, MeldHive, is the first commercial instantiation. The IP is the protective moat for both the product and any future licensees.
Filing posture
Active prosecution.
Provisional application filed March 18, 2026. Non-provisional in active drafting with target filing March 18, 2027. Foreign filing strategy under counsel review.
Our patent counsel is a top-tier US firm with deep AI/IP experience. The full filing strategy is shared under NDA.
Patent pending.
Licensing
What licensing looks like.
We are open to strategic licensing conversations with parties who want to build Engineered AI products in adjacent or non-overlapping verticals. We are not open to wholesale category transfer.
If you're a strategic acquirer, an enterprise platform, or a vertical-specific operator interested in the architecture rather than the product, contact us.
Discipline
Risk and discipline.
We operate with strict separation between public-facing material and patent-protected content. Mechanism details — how the architecture works internally — stay inside the patent and inside NDA-gated partner conversations. The public framing is outcome-language only: reliable, auditable, refuses-when-unsure, locally deployable.
This discipline protects the IP for licensing and protects foreign filing rights.
The pillars
Three layers of protection.
01
Patent pending
Lucid Decision LLC has filed with the USPTO and is operating under Patent Pending status. The filing covers the foundational architecture for high-accuracy AI in regulated decision environments — including the structural separation of how a question is understood from how an answer is verified.
02
Private implementation know-how
The patent protects the architecture. The methods used to optimize performance, govern validation behavior, and harden the system for real-world deployment remain inside the company.
03
Durable market position
As models improve and become easier to access, protected system design becomes more valuable, not less. Commodity access to language models lowers the barrier to building AI products, but it raises the value of structural differentiation at the architecture level, which is exactly what Engineered AI provides.
The moat
Why the moat matters.
Most AI moats erode quickly. Models improve, compute gets cheaper, and surface-level techniques spread fast.
Engineered AI's advantage sits higher in the stack. The moat is not data. It is not scale. It is not training spend. It is the protected architecture itself, which is exactly where defensibility matters in regulated markets.
Competitive context
What this is, and isn't.
Foundation models improve language generation. Retrieval improves factual grounding. Neither solves interpretation at the architectural level or produces reproducible, traceable outputs on its own.
Engineered AI works with those systems while adding capabilities they do not provide. That is what makes the advantage durable as the broader AI market matures.
Open a licensing conversation.
Strategic acquirers, enterprise platforms, vertical-specific operators, and patent counsel are welcome.