AI Geopolitical RiskAssessment for VC Diligence.
Country risk scores compress a nation into a single number. Institutional committees underwriting an asset need a structural read on where that specific company is physically, contractually, and regulatorily exposed. This guide describes how the Zurvek Engine surfaces geopolitical fragility from a pitch deck and data room, using an adversarial multi-agent pipeline and the Geopolitical Conflict Radar.
The Four Structural Signal Layers
Geopolitical risk, at the asset level, is not one variable. Four layers determine whether a company survives a corridor closure, an export-control shift, or a sanctions revaluation.
Supply Chain Concentration
Single-country component dependency, choke-point shipping lanes, and rare-earth or semiconductor exposure are surfaced as structural fragility, not as a rating.
Sector Concentration in Conflict Zones
Revenue, manufacturing footprint, and customer base are mapped against active and dormant conflict corridors so committees can see where value is physically located.
Counterparty & Sanctions Adjacency
Beneficial ownership, cross-border holding structures, and adjacency to sanctioned entities are traced through public registries rather than inferred.
Regulatory & Export-Control Drift
Shifts in export control regimes, dual-use classifications, and cross-border data rules are tracked as regime risk that can revalue an asset overnight.
How the Adversarial Pipeline Reads a Deck
A single language model summarising a pitch deck will smooth over geopolitical exposure because the deck itself omits it. Five agents, operating adversarially, refuse to let the omission stand. Each agent owns a slice of the structural read.
- FiscalHawk
Currency exposure, cross-border revenue mix, and capital control sensitivity under conflict-era FX regimes.
- MarketRadar
Demand fragility in sanctioned or contested markets, and re-routing cost when a corridor closes.
- RiskWarden
Supplier concentration, single points of failure, and the physical geography of production and delivery.
- Governor
Regulatory posture, export-control fit, and the institutional posture required to hold the asset through a shock.
- SenseCore Adjudicator
Resolves adversarial disagreement between agents and produces the committee-facing structural read.
The Geopolitical Conflict Radar
The Radar is not a rating surface. It plots the company's disclosed footprint against active and dormant conflict corridors, sanctions regimes, and export-control zones, and marks the intersections. A committee reads the Radar to see, in one frame, which corridors matter for this specific asset. Where the footprint is unknown, the Radar shows an explicit blind spot rather than filling it.
Five Questions the Output Answers for Committee
- Which single geography, if closed for 90 days, materially compresses this company's revenue?
- Where does the cap table intersect with sanctioned or politically exposed structures?
- What percentage of the input stack is sourced from a single conflict-adjacent supplier?
- How does the regulatory posture change if the target market's export regime tightens?
- What is the observable, sourced evidence for each of the above — not the founder narrative?
What the System Refuses to Do
By protocol, the Zurvek Engine does not forecast conflict onset, name winners, or price an asset. It surfaces structural exposure and cites the source of every claim. Probability, weighting, and the final decision remain with the committee. This is the same discipline applied across every risk dimension the platform reports on.
Questions
How is this different from a country risk score?
Country risk scores compress a nation into a single number. A structural read maps a specific company's revenue, suppliers, counterparties, and regulatory posture against active conflict corridors, and surfaces where that specific asset is fragile. The output is decision friction for a committee, not a rating.
Does this replace human diligence?
No. The adversarial pipeline compresses discovery so committees spend their time on judgement, not on assembling the underlying evidence. Every claim in the output cites its source, so a human diligence lead can verify and challenge every line.
How current is the conflict and sanctions data?
Public registries, sanctions lists, and export-control notices are re-queried at evaluation time rather than cached, so a verdict reflects the regulatory posture at the moment the committee received it.
What is intentionally excluded?
Predictions of conflict onset, timing, and outcome are excluded by protocol. The system surfaces structural exposure and lets the committee assign probability. This is the same discipline the platform applies to every other risk dimension.
Read the structural view on a real deck.
Upload a pitch deck and the adversarial pipeline returns a sourced, committee-grade read in under 180 seconds. Free tier covers the first evaluations.