Military operations floor with analysts at multi-screen command workstations
SynapseCommand
Sovereign Multi-Agent AI for Military Command

Decision Superiority for Military Command & Control.
On Your Terms.

SynapseCommand® augments command-and-control platforms with doctrinally grounded agents, each purpose-built for a staff task, and a natural-language commander interface: sovereign, post-quantum-secure, and auditable via hash-chained, append-only decision records.

Sovereign decision intelligence for NATO Allies, EU Member States and trusted partner nations
ORCHESTRATION RINGSOVEREIGN AGENT MESHN
Plain Language

What is SynapseCommand®?

SynapseCommand® is an AI platform that helps military commanders see faster, decide faster, and act with confidence. It plugs into existing command-and-control systems, runs on infrastructure you control, including air-gapped and classified environments, and writes every decision it supports to a hash-chained, append-only record that is ML-DSA-signed in assured deployments.

Schematic of the SynapseCommand agent mesh: a central command node linked to the agent nodes inside a dashed air-gap perimeter.
Watch on YouTubeRuntime 1:07 · English captions · full transcript below
Video transcript

Ninety seconds ago, this was a radar return. Now it is a decision.

Doctrinally grounded agents. One sovereign mesh. And no raw source data leaving your boundary without an authorised release policy.

Options ranked. Reasoning shown. Evidence sealed.

Then it stops. And waits.

The system recommends. It does not decide. A human does.

Your data. Your weights. Your jurisdiction.

Air-gapped. Post-quantum. Accountable to the last inference.

SynapseCommand. Decision superiority. On your terms.

01
Ingest

Pulls ISR, OSINT and force data from your existing systems via STANAG-aligned adapters.

02
Reason

Agents reason in parallel and coordinate through the sovereign Agent Mesh, with the combination set by mission requirements.

03
Decide

Commanders interact in natural language and receive ranked, traceable options, auditable via hash-chained, append-only decision records, with ML-DSA signing in assured deployments.

Decent Cybersecurity selected for NATO DIANA Decision Superiority Challenge
NATO DIANA · Decision Superiority Challenge

Selected for NATO DIANA.

Decent Cybersecurity is one of ten companies selected with SynapseCommand® for the NATO Defence Innovation Accelerator for the North Atlantic (DIANA) Decision Superiority Challenge, receiving €100,000 in contractual funding and working towards a live demonstration to NATO commands by December 2026. Selected in the NATO DIANA Decision Superiority Challenge, ranking in the top 2.34% of 427 submissions across 29 Allied nations.

Learn more on nato.int
1 of 10
Selected innovators
2.34%
Top of submissions
427
Total submissions
29
Allied nations
Built to connect. Not to replace.

Designed for existing command and mission environments

SynapseCommand® adds decision intelligence above the command, control, mission and operational-data environments you already run. Nothing is ripped out. Integration status varies by system and is stated per system in the compatibility catalogue, and the adapters we publish are a separate thing from any third-party system integration.

Keep your C2. Add the decision layer.

  1. 01
    Existing C2 and mission systems

    Stay in place, under your control.

  2. 02
    Authorised data exchange

    Defence standards and published adapters.

  3. 03
    SynapseCommand

    Classification-aware gateway and agent mesh.

  4. 04
    Decision intelligence

    Options, evidence and an auditable record.

Connect → Understand → Rehearse → Decide → Execute → Replay

C2 and mission systems

Integration pathways for NATO, national and vendor command environments. Each entry carries its own status and its own evidence. Names are used for identification only.

Open interoperability

Published adapters map external operational data into the SynapseCommand® canonical model before it reaches the decision layer. Implementation details, fixtures and tests are publicly inspectable. The integration contract is open; the decision runtime is not.

These are data formats, not third-party products. TAK here means the Cursor-on-Target message format; the TAK Server, ATAK and WinTAK applications are listed as systems above.

NATO SECRET cleared
EU SECRET cleared
NSPA registered supplier
NCIA Basic Ordering Agreement (BOA) holder
ESA registered
ISO 9001 / 27001
NSA-certified NIS2 auditor
Newsroom

Programme updates

Conceptual illustration of hybrid threats including military pressure, cyber operations, disinformation, economic coercion and critical infrastructure
Analysis

What Are Hybrid Threats? Cyber, Disinformation, Sabotage and Coercion Explained

Hybrid threats are coordinated harmful activities carried out with malign intent to undermine a target such as a state or institution through a variety, and often a combination, of means. They can include cyber operations, information manipulation, sabotage, economic coercion, covert political activity and military pressure. Hybrid campaigns often exploit vulnerabilities, ambiguity and difficulties of detection and attribution, frequently while operating below the threshold of open armed conflict.

Read update
Military Common Operational Picture showing air, land, maritime, space, cyber and information domains in a command centre
Analysis

What Is a Common Operational Picture (COP)? Military C2 Explained

A Common Operational Picture, or COP, is a shared and dynamic representation of operational information that can be tailored to support situational awareness, collaborative planning and decision-making. In military command and control, a COP brings relevant information from multiple sources into a shared, mission-focused picture. It is not simply a digital map, and it is not necessarily a single software application.

Read update
Warfighting Agent Suite

Doctrinally grounded agents. One sovereign mesh.

Each agent is purpose-built around a doctrinal task and governed by a fail-closed model register (an unapproved model tag is refused, never inferred), a five-role RBAC capability matrix with hash-only token storage, and a supervisor whose quarantine powers always carry an actor and a justification. Approval gates are enforced by the engine, not by a prompt: bounded, reversible automation runs human-on-the-loop, and every explicit approval gate is human-in-the-loop. Agents reason in parallel and coordinate through the sovereign Agent Mesh, with the combination set by mission requirements.

All agents complete the same verification programme. The tiers below record the technology readiness level at which each has been demonstrated.

TRL 7

System prototype demonstrated in an operational environment.

Capability in operational context.

From national command centres to disconnected edge nodes, SynapseCommand deploys where decisions are made.

Deployment Models

Where you need it. How you need it.

Secure sovereign data centre hall with rows of server racks

Sovereign Cloud

In-country, accredited cloud, under your jurisdiction and operational control.

Access-controlled door to a classified server enclave

On-Premises Classified

Designed for deployment up to NATO SECRET. Classification enforcement is implemented in code in both directions: unlabelled requests do not cross the gateway, answers above the caller's clearance are withheld, and withheld events are counted. Accreditation is per deployment.

Ruggedised tactical edge compute cases deployed in a field shelter

Air-Gapped Tactical

Runs disconnected. Inference, all operational agents and the decision record are local. Offline map packs are staged per node. The White Force Agent is training-variant only and excluded from operational builds.

Ready to see SynapseCommand® in action?

Briefings available under NDA. Classified briefings arranged through cleared channels.