Ten temporal dimensions.Eight predictive horizons.Total foresight.
Most intelligence systems report what has happened. The Sovereignty Infinium forecasts what will happen — from sub-second crisis emergence to generational strategic shift. Ten temporal dimensions run simultaneously. Eight predictive horizons, eleven model categories, and seven wargaming types feed a single, time-aware intelligence picture. Time is not a filter. Time is a dimension.
10 TEMPORAL DIMENSIONS
8 PREDICTIVE HORIZONS
11+ MODEL CATEGORIES
7 WARGAMING TYPES
SHERMAN KENT · BRIER
Time is the fourth dimension of intelligence.
Every signal has a temporal tag. Every entity has a temporal trajectory. Every claim has a half-life. Every prediction has a horizon.
Decisions operate on different clocks. A crisis manager needs sub-second alerts. A central bank needs a 6-month outlook. A defense planner needs a 10-year strategic assessment.
The platform does not ask the user to choose one clock and discard the rest. It runs all of them, simultaneously, against the same knowledge graph.
The platform's temporal architecture is the only one in the industry engineered from the ground up around the principle that time is a dimension, not a filter.
Time as a continuous variable
REALTIME
<1s
IMMEDIATE
1m–1h
OPERATIONAL
1–24h
TACTICAL
1–7d
CURRENT
1–4wk
STRATEGIC
1–6mo
MEDIUM
6mo–2y
LONG
2–10y
CHRONOS
10–25y
GENERATIONAL
25y+
Every signal is placed on this axis · every forecast is placed on this axis · every entity carries its trajectory on this axis.
The ten temporal dimensions.
The dimensions are not buckets. They are continuous. A claim made in REALTIME may be relevant in STRATEGIC; the platform maintains the link.
REALTIME
<1 second
Streaming ingest, alerting, network-anomaly detection
→ Sub-second crisis emergence, network intrusion
IMMEDIATE
1 minute – 1 hour
Active monitoring, dynamic baselines, fast-loop analysis
→ Breaking narrative, surge detection
OPERATIONAL
1 – 24 hours
Daily intelligence cycle, alert disposition, daily brief
→ Operational decision, sector watch
TACTICAL
1 – 7 days
Weekly tactical report, hunt hypothesis, campaign tracking
→ Tactical pivot, hunt plan
CURRENT
1 – 4 weeks
Monthly intelligence product, trend analysis
→ Current state, monthly board brief
STRATEGIC
1 – 6 months
Strategic assessment, scenario modeling, estimative product
→ Strategic posture, mid-horizon decision
MEDIUM
6 months – 2 years
Trend and regime-shift detection, market trajectory
→ Medium-horizon policy, capital allocation
LONG
2 – 10 years
Long-cycle foresight, technological trajectory, climate shift
→ Long-horizon investment, generational policy
CHRONOS
10 – 25 years
Civilizational foresight, demographic transition, secular trend
→ Decadal posture, alliance planning
GENERATIONAL
25+ years
Multi-generational foresight, climate trajectory, civilizational shift
→ Generational institution design
The eight predictive horizons.
Each horizon has a model class, a confidence expectation, a refresh cadence, and a decision context.
Flash
<1 hour
Streaming anomaly · network burst
Low–Medium (fast)
Crisis manager · SOC · watch officer
Near
1–24 hours
Time-series forecast · pattern-of-life deviation
Medium
Sector operator · daily brief
Short
1–7 days
Tactical predictive · hunt-hypothesis
Medium
Hunt team · tactical lead
Medium
1–4 weeks
Trend extrapolation · regime-shift
Medium–High
Operations lead · monthly board
Strategic
1–6 months
Scenario modeling · war-gaming · Delphi
High (Sherman Kent)
Strategic decision · policy
Long-cycle
6 months – 2 years
Trend · structural · market trajectory
High
Capital allocation · mid-horizon policy
Decadal
2–10 years
Long-cycle foresight · technology · climate
Medium (structural)
Investment · alliance · posture
Generational
10+ years
Civilizational · demographic · secular
Low–Medium (range)
Generational institution design
Why eight, not five or twelve. Eight horizons reflect a deliberate balance. Five would conflate decision tempos that should be distinct. Twelve would fragment the model class and erode confidence. Eight matches the cadence at which a sovereign decision-maker actually consumes intelligence.
The eleven model categories.
The platform does not bet on a single modeling paradigm. It runs eleven model categories in parallel, with explicit weighting and ensemble logic.
Time-series forecasting
Flash → Short
Strengths
Pattern extrapolation · seasonality
Limits
Regime shifts break it
Anomaly detection (statistical)
Flash → Short
Strengths
Catches what has never been seen
Limits
High false-positive rate without tuning
Anomaly detection (ML-based)
Flash → Medium
Strengths
Multivariate · contextual
Limits
Requires labeled data
Bayesian inference
Short → Long
Strengths
Explicit priors · updates with evidence
Limits
Priors must be set carefully
Agent-based modeling
Medium → Strategic
Strengths
Captures emergent behavior
Limits
Calibration is hard
System dynamics
Medium → Decadal
Strengths
Feedback loops · structural
Limits
Aggregate only
Scenario modeling
Strategic → Long
Strengths
Handles deep uncertainty
Limits
Subject to scenario-design bias
Delphi consensus
Strategic → Decadal
Strengths
Aggregates expert judgment
Limits
Subject to panel composition
War-gaming (red/blue)
Tactical → Strategic
Strengths
Adversarial dynamics · intent
Limits
Resource-intensive · depends on play quality
Causal inference
Medium → Decadal
Strengths
Distinguishes correlation from causation
Limits
Counterfactual identification is hard
Graph neural network
Flash → Long
Strengths
Network dynamics · propagation
Limits
Interpretability is a challenge
The seven wargaming types.
Wargaming is the platform's stress-test for predictive products. The platform does not treat wargaming as a single activity; it treats it as a typology of seven distinct forms.
01 / 07
Strategic
- Players
- 2–4 teams · senior decision-makers
- Tempo
- Months per move
- Context
- National posture · alliance decisions
02 / 07
Operational
- Players
- 2–6 teams · mid-level planners
- Tempo
- Days per move
- Context
- Campaign planning · crisis response
03 / 07
Tactical
- Players
- 2 teams · operators
- Tempo
- Hours per move
- Context
- Engagement-level decisions
04 / 07
Cyber
- Players
- 2 teams · technical operators
- Tempo
- Variable
- Context
- Adversary emulation · defense validation
05 / 07
Cognitive
- Players
- 2 teams · narrative operators
- Tempo
- Weeks per move
- Context
- Influence operations · narrative warfare
06 / 07
Economic
- Players
- 2–4 teams · economic actors
- Tempo
- Quarters per move
- Context
- Sanctions · trade · market intervention
07 / 07
Political
- Players
- 2–4 teams · political actors
- Tempo
- Months per move
- Context
- Coalition · election · diplomatic posture
Each wargame produces an After-Action Report (AAR) that feeds the calibration loop. The platform tracks which moves were expected, which surprised, and which hypotheses were disconfirmed. The result: each wargame sharpens the next.
Confidence, calibration, and the limits of foresight.
The platform is candid about the limits of prediction. Three calibration mechanisms sit at the core of the temporal surface.
Sherman Kent scale
Estimative probability reported on Sherman Kent conventions: Low (≈40%), Medium (≈60%), High (≈80%), with a structured analytic chain. Probabilities are not the point; the chain is.
Brier score
Every predictive model is scored against ground truth using the Brier score. Models are reweighted based on their rolling Brier performance. Models that fail to calibrate are demoted or retired.
Adversarial validation
Every horizon is stress-tested with adversarial validation. The platform generates a synthetic future that disagrees with the forecast, and the forecast must defend itself against that future.
Calibration dashboard
Anonymized example — rolling Brier by horizon.
The platform does not defend a model on the basis of its elegance; it defends a model on the basis of its calibration.
| Horizon | Model | Rolling Brier | Calibration trend | Weight in ensemble |
|---|---|---|---|---|
| Flash | Anomaly-ML | 0.142 | Improving | 18% |
| Near | Time-series | 0.118 | Stable | 22% |
| Short | Hunt-hyp | 0.176 | Improving | 14% |
| Strategic | Scenario + Delphi | 0.221 | Stable | 31% |
| Decadal | System dynamics | 0.298 | Stable | 15% |
AI + human tradecraft across the temporal dimensions.
At REALTIME and IMMEDIATE, AI leads. At STRATEGIC and beyond, humans lead. The middle is shared. The handoff is explicit.
REALTIME
Streaming detection, alert routing, threshold management
Watch officer judgment on disposition
Sub-second
IMMEDIATE
Fast-loop analysis, pattern-of-life deviation
Operator judgment on escalation
Minutes
OPERATIONAL
Daily cycle, dossier generation, alert disposition
Analyst review, bias check
Hours
TACTICAL
Hunt hypothesis, TTP mapping, trend extrapolation
Hunt-team validation
Daily
CURRENT
Trend analysis, monthly product generation
Analyst authoring, editorial review
Weekly
STRATEGIC
Scenario model run, Delphi aggregation
Estimative product authorship
Monthly
MEDIUM
Bayesian inference, structural model
Analyst review, alternative scenarios
Quarterly
LONG
System dynamics, trajectory model
Foresight-team review, institutional review
Quarterly
CHRONOS
Civilizational data analysis
Generational thinking, policy review
Annual
GENERATIONAL
Climate, demographic, structural
Civilizational judgment, ethical review
Multi-year
The axis along which everything is parameterized.
Temporal intelligence is the dimension along which the rest of the platform is parameterized. Every capability carries a temporal tag.
Multi-INT Fusion — Each INT signal has a temporal tag. The fusion engine resolves entities across time as well as across INTs.
Real-Time Crisis Intelligence — Crisis workflows operate on the REALTIME → IMMEDIATE → OPERATIONAL axis.
Reputation & Perception — 90+ sub-metrics are computed at multiple temporal resolutions.
Threat Detection & Attribution — Threat-actor dossiers maintain temporal trajectories.
Disinformation & Influence Operations — The campaign-anatomy model is a temporal model.
Media Intelligence — Volume, velocity, reach, virality — all temporal metrics.
Geopolitical Foresight — Country risk forecasting and bilateral posture tracking are scored on the temporal axis.
AI & LLM Perception — Perception drift over time across CHRONOS.
Cyber Threat Intelligence — IOC aging, TTP evolution, vulnerability lifecycle — all temporal.
Financial & Economic Intelligence — Market microstructure, sovereign credit, sanctions-evasion network evolution — all temporal.
Geospatial & Physical Intelligence — Pattern-of-life, change detection, route reconstruction — all temporal.
Command Center — Display-wall presets map to temporal contexts.
Outcomes that are time-specific.
6-week
Lead on coordinated inauthentic behavior
CHRONOS-aware detection · REALTIME alerting
4-day
Lead on cyber staging operation
TACTICAL forecast · OPERATIONAL handoff
6-month
Lead on market-structure inflection
STRATEGIC scenario · MEDIUM forecast
24-hour
Lead on narrative surge
IMMEDIATE fast-loop · REALTIME escalation
Decadal
View on technology transition
LONG model · CHRONOS review