Generate. Enrich. Decide.
From 'I have seen something' to 'the system knows what it is, where it fits, and what should happen next.'
Every targeting failure starts the same way: a real detection, reported messily, cleaned up too late. The sensors are delivering. The infrastructure is being built. The missing layer turns messy reporting into structured data, structured data into contextualised intelligence, and intelligence into options a commander can weigh and sign. Speed without data quality produces errors at machine speed. The Intellium Targeting Suite is that layer.
The mission thread
ISR, counter-fires, counter-UAS and partner feeds detect.
A structured workflow validates at the point of capture: fields complete, grid sanity-checked against bearing and range, equipment confirmed against authoritative naming, impossible combinations flagged before dissemination.
Task organisation, NATO designations and symbology, geospatial context, high-payoff and high-value tags, system-of-systems relationships.
Source reliability and information credibility graded separately, the way analysts are trained to judge. Gaps held and chased, not papered over. Every action logged.
Intelligence publishes to the picture and tools in use. Where options are the need, AiPower generates and argues courses of action against the commander's guidance, every quantitative claim verified by deterministic code.
Structured battle damage assessment feeds the collection plan, re-tasking and target confidence.
AiSource
Generate with Structure
AI-ready target data, generated and validated at the source. Structured Sighting, BDA, HLR, GMTI and counter-UAS reporting, with validation built into the form, so raw observation becomes machine-readable target data before it touches the network.
AiWorkbench
Enrich with Context
Enriched, quality-gated target intelligence, exported to where decisions happen. It ingests structured reports, chat traffic and the feeds you choose to expose, extracts what can be trusted, grades it, and routes it onward or asks for what is missing.
AiPower
Decide with Agentic AI
Intellium's agentic prototype for course-of-action generation under command. AI agents draft and argue options; deterministic code verifies every claim; a human signs the decision. A full three-option estimate in minutes, not days.
Proven where targeting data actually moves.
Every figure below was measured on exercise and is shared with the agreement of the formations involved. The full record is available in a capability briefing.
CORPS-LEVEL EXERCISE, 2026
20 chat rooms; 7,032 messages, 2,715 of them unstructured.
234 targets surfaced that manual triage missed.
59,061 targets enriched and published to the live picture.
Triage time down by more than half.
95% detection against the high-payoff target list.
Live integration with the in-service battle management application.
Where the suite has been exercised
Greyhound Cyclone 26, Exercise ARRCADE STRIKE, the Warfighter Reloaded divisional exercise, and the ASGARD Decide integration event.
Deployed across classifications from OFFICIAL SENSITIVE to OpNET Red.
These results belong to AiSource and AiWorkbench, the two products proven on exercise. AiPower is a prototype in development.
Built to power the ecosystem you already run, not replace it.
AiWorkbench does not replace your CIP or COP; it powers them. Structured, enriched, auditable target data flows out to the systems already in service: SitaWare, ArcGIS, ComBAT, DTW, ASGARD and Zodiac, over open interfaces and standard message routes.
Built for the force being built
The direction of travel is public: a Digital Targeting Web connecting sensors, deciders and effectors; ASGARD accelerating land targeting; the Land ISTAR Programme modernising find-and-understand; a Defence AI taskforce pushing AI into planning. The suite is engineered for that world: hub-and-spoke around a customer-owned source of truth, open interfaces at every boundary, containerised from tactical edge to cloud, models as replaceable components. Integrate once; connect new sensors and consumers without re-architecting.
Exploitable, composable, sovereign.
Exploitable by design
API-first and event-driven, with an OpenAPI specification, loosely coupled services, no proprietary data formats, and federated access and authorisation. Secure by design.
NATO data standards and interoperability
STANAG 5658 targeting conceptual data model, AJP-3.9 joint targeting cycle, APP-6 symbology and SIDC enrichment, APP-11 (ADatP-3) messaging, NATO C3 taxonomy alignment, and DTW Targeting API conformance.
Programme conformance
Conforms to the ontologies and schemas of the programmes already in the plan: Palantir Ontology for ASGARD, Zodiac SDL for Land ISTAR, and MOD Schema over XMPP and chat. Intellium's data arrives in the shape each programme already expects. Designed to be adopted within existing programme structures, not around them.
Data governance and assurance
A full audit trail: every message ingested, assessed or discarded is logged. Provenance per entity: source, time, classification and confidence. Configurable compliance rules, versioned and enforced at ingest. Analyst-in-the-loop: AI recommends, a human confirms, the decision is recorded.
Sovereign UK. Aligned with NATO.
100% British-owned IP. Open formats and published interfaces mean a boundable cost of exit: no lock-in. A sovereign UK company, built to NATO standards and proven alongside NATO formations, including on Exercise ARRCADE STRIKE.