Use Cases

    Where decision superiority matters most.

    Five representative mission profiles, each showing what goes in, which modules reason over it, and what the commander receives. Every schematic below is illustrative; the deployed pipeline is configured to your sensors, doctrine and accreditation boundary.

    MaritimeMulti-domainCoalitionCounter-UASPredictive
    UC-01

    Maritime & subsea-infrastructure protection

    Continuous fusion of radar, AIS, SAR, OSINT and SIGINT to detect anomalous UAV and vessel activity around critical maritime and subsea infrastructure. SynapseCommand® generates ranked, traceable courses of action and seals the evidence chain for downstream legal and operational use.

    Offshore patrol vessel on station near an offshore wind farm in low visibility
    Maritime domain · patrol asset on station near critical subsea infrastructure
    12
    Operational Agents

    Built, versioned and independently removable; the training-only White Force module is excluded from operational builds

    100%
    Decision audit coverage

    Design property: every decision event is written to a hash-chained, append-only record, with ML-DSA signing in assured deployments

    SENSOR FUSION → COA → SEALED EVIDENCEMODE: ILLUSTRATIVEFRIENDLY TRACKANOMALYUNCUED THREATINGESTRADAR / OTHAIS · VMSSAR TASKINGMARITIME OSINTSIGINT CUEFUSIONTRACK GRAPH5 SRC · 1 PICTURERANKED COURSES OF ACTIONCOA-1 SHADOWP 0.62COA-2 QUERYP 0.27COA-3 INTERDICTP 0.11SIGNED LOGSHA-384 · PQ SEALAREA OF INTEREST · SUBSEA CABLE CORRIDORFRIENDLY 04AIS GAP 41 MINUNCUED UAV
    Fig. 1: illustrative schematic, not operational data
    Inputs
    • Coastal & OTH radar
    • AIS / VMS feeds
    • SAR imagery tasking
    • Maritime OSINT
    • SIGINT cueing
    Modules engaged

    Registered agents carry an SC- identifier; other entries are platform services or functions within an agent.

    SC-ISR-01 ISR Fusion AgentSC-COA-01 COA Analysis EngineThreat-assessment functionProvenance and audit service
    Commander receives
    • Anomaly dossier with confidence bounds and contributing sensors
    • Three ranked COAs with legal and escalation notes
    • Evidence bundle designed to support evidential integrity, chain-of-custody and legal review; admissibility remains jurisdiction- and procedure-dependent
    UC-02

    Multi-domain wargaming & COA analysis

    Red and blue agents simulate adversary doctrine and friendly posture across land, maritime, air, space and cyber. The COA Analysis Engine explores high-dimensional outcome spaces with reinforcement learning, Bayesian inference and Monte Carlo methods.

    Staff officers working over map sheets and laptops in a planning room
    Planning cell · red and blue staff work the branch-and-sequel problem
    256
    Branches per scenario

    Customer-configurable; default sampling depth in the reference configuration

    1,685
    Verified question-runs

    One file per agent per model set, each with commit, versions, models and per-case pass rates

    COA BRANCH EXPLORATION · MONTE CARLO8 400 RUNS / HFAVOURABLECONTESTEDADVERSEDOMAINSLANDMARITIMEAIRSPACECYBERREDBLUEADVERSARIAL SELF-PLAYH-HOURB1B2B3B4B5BRANCHSEQUELOUTCOME SPACEMISSION SUCCESS72%ATTRITION RISK41%ESCALATION18%LOGISTICS FAIL29%C2 DEGRADATION23%
    Fig. 2: illustrative schematic, not operational data
    Inputs
    • Order of battle
    • Doctrinal templates
    • Terrain & METOC
    • Logistics state
    • Cyber & EW posture
    Modules engaged

    Registered agents carry an SC- identifier; other entries are platform services or functions within an agent.

    SC-RED-01 Red-Force Behaviour AgentSC-BLUE-01 Blue-Force Posture AgentSC-COA-01 COA Analysis EngineSC-STR-01 Strategy & Doctrine Agent
    Commander receives
    • Branch-and-sequel tree with probability-weighted outcomes
    • Decision points, triggers and commander's critical information requirements
    • Sensitivity analysis showing which assumptions move the result
    UC-03

    Coalition operations across allied nations

    Allied nodes federate reasoning across the Post-Quantum Coalition Fabric. In the reference federated deployment, source records remain at the originating node. Only policy-authorised queries, derived outputs and required provenance metadata cross the federation boundary. Exceptions and configuration choices, including what derived artefacts (e.g. embeddings, model outputs, confidence values) may cross the boundary, are documented per deployment. Policy-based output filtering and release controls apply to each output, tested against semantic and indirect information leakage; provenance is preserved across the whole coalition.

    Officers seated in a darkened coalition headquarters briefing theatre
    Coalition headquarters · multinational staff, national data held locally
    0
    External AI calls in runtime code

    Enforced by a test that parses every file

    32
    Allied nodes supported

    Customer-configurable; tested federation size in a representative test environment

    100%
    Release-control coverage

    Design property: every output passes policy-based release filtering before disclosure

    FEDERATED REASONING · SOURCE RECORDS STAY LOCALPQ FABRIC · ACTIVENATIONAL NODEGRADIENT EXCHANGEPOLICY GATEPQ COALITIONFABRICML-KEM · ML-DSACZEWEIGHTS STAY LOCALSVKWEIGHTS STAY LOCALPOLWEIGHTS STAY LOCALAUTWEIGHTS STAY LOCALIRLWEIGHTS STAY LOCALGRADIENTS ONLYRECORDS STAY LOCALTOKEN-LEVEL GATECAVEAT-AWARE
    Fig. 3: illustrative schematic, not operational data
    Inputs
    • National enclaves
    • Releasability caveats
    • Coalition PKI
    • Shared mission graph
    Modules engaged

    Registered agents carry an SC- identifier; other entries are platform services or functions within an agent.

    SC-PLAN-01 Operational Planning AgentFederation broker servicePolicy enforcement serviceProvenance and audit service
    Commander receives
    • Shared operating picture assembled from local inference only
    • Per-token releasability decisions, individually attributable
    • Coalition-wide provenance chain surviving node disconnection
    UC-04

    Counter-UAS & short-range air defence

    Low, slow and small threats compress the decision cycle to seconds. SynapseCommand® correlates passive RF, EO/IR and radar into a single track picture, applies rules of engagement as executable policy, and presents the operator with a pre-validated response recommendation, human-in-the-loop: engagement requires explicit approval by a qualified operator, after which the approved recommendation is passed to the host fire-control system, which tasks the effector. SynapseCommand does not fire or command an effector directly.

    Ground-based air defence radar and short-range effector vehicle at dusk
    SHORAD site · sensors and effectors at the compressed end of the decision cycle
    708
    Platform tests

    Plus 484 engine and conformance tests and 404 offline tool-and-expectation checks

    100%
    ROE checks automated

    Design property: every engagement recommendation passes the configured policy set; operator approval remains mandatory

    COUNTER-UAS · DETECT TO EFFECTHUMAN ON LOOPCONFIRMED TRACKCLUTTER / UNKNOWNSENSOR RETURNTRK-071PASSIVE RF + EO/IR + SHORAD RADARSENSOR SCOPE · 6 KMKILL-CHAIN TIMELINEDETECTT+0.0s01CLASSIFYT+0.4s02CORRELATE TRACKST+1.1s03ROE CHECK · POLICYT+2.3s04OPERATOR DECISIONT+3.0s05EFFECTOR TASKEDT+3.4s06CONFIDENCERF91%EO/IR78%RADAR66%FUSED96%
    Fig. 4: illustrative schematic, not operational data
    Inputs
    • Passive RF detection
    • EO/IR trackers
    • SHORAD radar
    • Airspace control order
    • ROE policy set
    Modules engaged

    Registered agents carry an SC- identifier; other entries are platform services or functions within an agent.

    SC-ISR-01 ISR Fusion AgentSC-TGT-01 Targeting & Engagement AgentThreat-assessment functionPolicy enforcement service
    Commander receives
    • De-duplicated track with classification confidence per sensor
    • ROE-checked engagement recommendation and alternatives
    • Sub-four-second detect-to-decision timeline, fully logged
    UC-05

    Predictive sustainment & logistics resilience

    Sustainment failure ends operations more often than contact does. The platform forecasts Class I–IX consumption against the current scheme of manoeuvre, scores convoy and route risk, and proposes re-allocations before the shortfall becomes a culminating point.

    Military logistics depot with palletised supplies and materiel handling equipment
    Sustainment node · Class I-IX stocks against the scheme of manoeuvre
    72 h
    Shortfall warning lead time

    Customer-configurable forecast horizon in the reference configuration

    100%
    Convoy risk coverage

    Design property: every planned route is scored before it is presented

    PREDICTIVE SUSTAINMENT · CLASS I–IXHORIZON 72 HNOMINALAT RISKCRITICALDISTRIBUTION NETWORKDEPOTNODE 01RAIL HEADNODE 02FWD BASENODE 03UNITNODE 04PREDICTED SHORTFALL · 72 HCONF. BAND ±6%CLASS III FUEL62%CLASS V AMMO84%CLASS VIII MED31%CLASS IX PARTS47%RE-ROUTE PROPOSEDCONVOY RISK SCOREDTRACEABLE SUSTAINMENT LOG
    Fig. 5: illustrative schematic, not operational data
    Inputs
    • Consumption telemetry
    • Convoy & route data
    • Maintenance state
    • Scheme of manoeuvre
    • Host-nation support
    Modules engaged

    Registered agents carry an SC- identifier; other entries are platform services or functions within an agent.

    SC-LOG-01 Logistics AgentSC-COA-01 COA Analysis EngineRoute-risk functionProvenance and audit service
    Commander receives
    • 72-hour shortfall forecast by supply class with confidence bands
    • Re-routing and cross-levelling options ranked by operational impact
    • Traceable sustainment record, with ML-DSA signing in assured deployments, for audit and coalition reconciliation
    Staff officers working over a digital planning map table in a darkened operations room
    Coalition planning cell · illustrative
    Live simulation

    Full kill chain, end to end.

    Find, Fix, Track, Target, Engage, Assess. Watch SynapseCommand® run the complete chain with every agent hand-off, rules-of-engagement gate and audit event exposed. Illustrative data only.

    Low-slow-small incursion over a protected nodeDOMAIN AIR · SHORADTRACK TRK-071ROE-A2 · SELF-DEFENCE
    T+0T+0.0sT+4.6s
    F1FIND
    T+0.0s
    Passive RF anomaly at 5.8 km, bearing 043
    SC-ISR-01 ISR Fusion Agent
    F2FIX
    T+0.4s
    EO/IR slew cued; position fixed to ±11 m
    SC-ISR-01 ISR Fusion Agent
    T1TRACK
    T+1.1s
    Threat track correlated from three sensor returns
    SC-ISR-01 ISR Fusion Agent
    T2TARGET
    T+2.3s
    Class-1 UAS, hostile intent p=0.94; recommended response generated
    SC-ISR-01 threat-assessment function
    E1ENGAGE
    T+3.0s
    ROE gate passed; human-in-the-loop authority requested
    Policy enforcement service
    E2APPROVE
    T+3.2s
    Operator approval received; authority record verified in the hash-chained decision log
    Duty commander · human-in-the-loop
    E3HANDOFF
    T+3.4s
    Approved recommendation passed to host fire-control system
    SC-TGT-01 effects-planning function
    E4EXECUTE
    T+3.8s
    Host fire-control system tasks authorised RF effector
    Host fire-control system (external)
    A1ASSESS
    T+4.6s
    Effect assessment returned; decision chain sealed to ledger
    Provenance and audit service
    PROTECTED NODERFEO/IRRADARTRK-071PHASE F1 · FINDT+0.0s
    PASSIVE RF91%
    EO/IR0%
    SHORAD RADAR0%
    FUSED0%
    Decision log1/14 events
    T+0.0sRF-COLLECT: unclassified emitter, 2.4 GHz hopping
    Effector
    RF DEFEAT (NON-KINETIC) · TASKED BY HOST FIRE-CONTROL SYSTEM
    Outcome
    Chain in progress…
    Audit
    Recording every agent hand-off…

    Your mission profile isn't listed?

    SynapseCommand® is configured per deployment. Bring your sensors, doctrine and accreditation constraints and we will map them to the agent mesh under NDA.

    Contact