Galahad · Journal
Journal
Notes, methods, and points of view from the Galahad team.
-
AI deployments that will not fail in 2026
95% of generative AI pilots return nothing measurable. The problem is not the model, it is rented intelligence: why an ontology you own is the only AI asset that stays yours.
-
Getting your money's worth
Consultancies, labs and software vendors bill hours, tokens and seats: three ways to charge at the door. Why Galahad bills only on the value created, payable on production.
-
Decision systems for defence and aerospace
How we model defence and aerospace operations as an ontology so a readiness, sustainment or targeting question resolves once, under permissions the database enforces, and can be replayed a year later.
-
Decision systems for grid operations
How to model energy and utilities operations so an operator can answer a time-bound question across SCADA, GIS, OMS, EAM and settlement, and act on the answer under audit.
-
Modelling decisions in banking and insurance
How we model decision systems for banking, insurance and financial crime operations: what to declare first, what becomes an ontology object, how to handle systems you cannot write to, and why determinism and replay are the only way this survives model risk governance.
-
Decision systems for healthcare operations
How to model hospital and life sciences operations so a decision can be made and defended: what you declare first, what becomes an object, how purpose limitation is enforced at the query rather than in a prompt, and where the approach is the wrong tool.
-
Decision systems for manufacturing operations
How we model industrial manufacturing operations so that a containment, a scrap or a release decision can be answered across MES, ERP, PLM, quality and historian data, and acted on under control.
-
Public administration decisions that survive appeal
How to build a decision system across administrative registries when a citizen can contest the result: what you declare first, why warehouses and chat layers fail here, and what determinism, replay and the audit trail actually have to produce.
-
Security and due diligence answers
The questions security, audit and procurement functions are told to ask an agentic AI vendor, each answered with the mechanism that enforces it, the configuration it depends on, and the boundary where it stops.
-
Own the decision space: the case for sovereign, governed AI
Why the next advantage for governments and enterprises is sovereign, governed agentic AI: decision intelligence that reads across every data silo and acts under human control, on-premise, with a full audit trail.