- 01Entities ranked by exposure score
- 02Cross-reference to export-control licences and normative atoms
- 03Flags for non-EU ownership, single-source dependency and eligibility gap
- 04Source citation per element
Defence Finance Monitor is a structured intelligence infrastructure for the European defence and dual-use ecosystem. It renders strategic priorities, industrial capability, regulatory instruments, funding programmes and market signals within one semantic environment — where heterogeneous institutional material becomes comparable, retrievable and analytically ranked.
Retrieval, classification and ranking are governed by a closed ontology and a controlled vocabulary. The access layer is natural language; the system beneath is constrained, source-grounded and domain-specific.

Its function is to return evidentially grounded, ranked responses to precisely formulated institutional questions about the European defence and dual-use domain.
Institutional users interrogate the system on specific analytical questions — entity exposure, supply-chain composition, eligibility conditions, capability alignment, regulatory configuration, ownership structure, funding-programme relevance. The system resolves each question and returns a structured response.
Coverage extends across the European defence industrial base and the instruments that govern it: strategic priorities, operational and tactical capabilities, technology clusters, regulatory regimes, procurement records, funding programmes, ownership configurations and trade signals.
Every element of a response is traceable to an institutional source — a normative atom, a procurement record, a corporate filing, a strategic document, a trade signal. Rankings are deterministic under identical inputs. No fluency is introduced outside what the sources support.
The semantic layer is what makes heterogeneous institutional material analytically commensurable. Four components define it.
A formal model of the domain. Thirteen strategic priority codes, fifty-three operational nodes, fifty-two tactical capabilities and twenty-seven technology clusters define the admissible entities and the relations the system recognises. The ontology is bounded: nothing is classified outside it.
A fixed classification of defence objects: entities, capability areas, technology domains, regulatory regimes, procurement instruments and funding programmes. Ingested records are normalised against the taxonomy before they reach the analytical layer.
A controlled lexicon for defence and dual-use terminology. Synonyms, acronyms, institutional references and domain-specific expressions resolve to canonical forms. Meaning is retrieved, not inferred.
Seven certified layers feed the framework — Normative, Strategic Documents, Entity and Company Data, Intelligence Reports, Research, Prime Contractor Data, Market Signals. Sources are pre-classified before any query is processed: 27,440 atomic normative units, 1,802,619 trade-signal records and 26,284 export-control licences.
The analytical logic of the system rests on a specific proposition: strategic relevance, once recognised by EU and NATO instruments, becomes — over time — industrial, financial and regulatory relevance. The system is constructed to observe this translation at the point where it occurs.
Capability plans, regulatory frameworks, procurement programmes and funding instruments are treated as structured objects. They define the capability areas, the technological domains and the compliance conditions that shape the operational landscape.
From these signals the system derives operational consequences: which industrial nodes are structurally critical, which capability gaps carry funding implication, which regulatory regimes determine eligibility, which ownership and trade configurations become material. Analytical relevance is made explicit and traceable to the instrument from which it originates.
The access layer is natural language — a question phrased as the analyst would phrase it. What operates beneath is constrained: resolution against the closed vocabulary, structured retrieval across the seven knowledge layers, ranking by proprietary logic, and return of a source-grounded response. The cases below illustrate the analytical form of questions the system resolves and the structure of the responses produced.

Each of the five institutional profiles below operates on the same semantic architecture, applied to a distinct analytical agenda.