Confidential - Design Partner Preview

Percolation · PEI (Pressure-to-Equilibrium Intelligence)

Use Case Demo: Port Aurelia Chokepoint Crisis

A technical walkthrough of the scenario intelligence platform.

Designed for a live-demo or clickable prototype environment, this page walks a design partner, enterprise client, or internal stakeholder through the exact Port Aurelia scenario step by step.

Request design partner walkthrough Jump to intervention comparator

Date

[Date]

Presenter

[Name]

Audience

[Design Partner Name]

Slide 01

Live-demo walkthrough

Agenda

Platform Overview

5 min

Live Scenario Walkthrough: Port Aurelia Crisis

15 min

Technical Architecture and Integration

5 min

Q&A / Next Steps

Open

Presentation map

02The World Before PEI 03Operator Console 04Define the Shock 05Simulation Execution 06Breakpoint Identification 07Node Deep Dive

01 Title Slide 02 The World Before PEI 03 Operator Console 04 Define the Shock 05 Simulation Execution 06 Breakpoint Identification 07 Node Deep Dive 08 Burden Transfer Map 09 Intervention Comparator 10 Post-Intervention Path 11 Innovation Vector 12 Backtesting and Calibration 13 Technical Integration 14 What You Have Seen 15 Design Partner Engagement 16 Q&A and Appendix

Slide 02

The World Before PEI

You have data. You lack the mechanism of transmission.

Standard risk stacks can flag that a chokepoint is under pressure. They still stop short of showing which node breaks first, where losses concentrate, and which intervention changes the path.

Mock dashboard

Red alert

Alert state

Geopolitical Tension: Port Aurelia

Elevated risk in the Aurelia Strait.

Shipping futures up 15%.

News headlines: "Naval exercises near critical trade route."

What standard tools tell you

Elevated risk in Aurelia Strait.

Shipping futures up 15%.

News headlines: "Naval exercises near critical trade route."

What standard tools do not tell you

Which specific node in your portfolio or exposure breaks first?

How will insurers reprice risk in response?

Who absorbs the first dollar of loss?

What intervention actually changes the outcome?

The gap

We see the storm. We cannot see the damage path.

Cited basis

[1]World Bank (2025): Global trade growth is running at the slowest pace in a decade.

[2]Lloyd's (2024): Trade-route conflict scenarios imply up to $14.5T in global economic loss.

Slide 03

PEI Operator Console

This is the PEI Operator Console.

The operator lands on a pressure-native control surface: system health, active signals, scenario launch, and a preview of burden transfer are visible in one place.

Demo action

Click "Scenario Launcher" -> Select "Port Aurelia Closure - 14-Day Simulation."

World-State Health Index

78/100 (Elevated)

Active Pressure Signals

Port Aurelia Tension: 92% Breakpoint Probability

Scenario launcher

Port Aurelia Closure - 14-Day Simulation

Burden Transfer Map preview shows early yellow edges around the chokepoint before the cascade is fully visible.

A. World-State Health Index

Aggregate fragility score of the entire modeled system. Current state: 78/100.

B. Active Pressure Signals

Real-time feeds flagging rising tension, including Port Aurelia at 92% breakpoint probability.

C. Scenario Launcher

Dropdown to select pre-built scenarios or create a custom shock.

D. Burden Transfer Map Preview

Miniature force graph showing a mostly stable system with early yellow edges around the chokepoint.

Slide 04

Step 1 - Define the Shock

Define the pressure vector.

The scenario configurator turns a geopolitical headline into a structured pressure vector that the model can propagate through the network.

Demo action

Click "Run Scenario."

Shock Type

Chokepoint Closure

Location

Port Aurelia

Duration

14 days (initial estimate)

Severity

Complete blockage (no traffic)

Include Secondary Effects

Insurance repricing, government response, rerouting behavior

World-State Model (NetworkX/Neo4j) is queried: find all nodes dependent on Port Aurelia within 3 hops.

Predictive Layer (PyMC) is primed with prior probabilities based on historical chokepoint events.

Agent Policy Layer is loaded with institutional actors: 3 insurers, 50 importers, 200 suppliers, and 2 governments.

Cited basis

[3]UNCTAD (2024): Strategic maritime chokepoints remain one of the largest systemic threats to trade-dependent economies.

Slide 05

Step 2 - Simulation Execution

Running 10,000 parallel futures.

Once launched, the system fans out thousands of slightly varied futures across the compute layer, runs the full agent-based model, and streams early decisions while the scenario is still executing.

Demo action

Simulation completes. System prompts: "View Results."

Execution status

Simulating T=0 to T=90 days...

Monte Carlo paths

10,000

Time horizon

T=0 to T=90 days

Execution fabric

Ray cluster + Mesa/LangGraph agents

Live log stream

[Ray Worker 42] T+2: InsurerAgent 'GlobalRe' -> Action: Reprice War Risk (+300%)
[Ray Worker 17] T+2: ImporterAgent 'ElectroCorp' -> Action: Absorb Margin (Buffer 2% remaining)
[Ray Worker 88] T+5: GovernmentAgent 'Ministry of Trade' -> State: Monitoring (No trigger yet)
[Ray Worker 03] T+7: LogisticsNode 'Midwest Logistics Inc' -> Warning: Cash buffer < 5 days.

Ray distributes 10,000 slightly varied simulations across the cluster.

Each simulation runs the full Mesa agent-based model with LangGraph agents making constrained decisions.

Results are aggregated in real time so the operator can move directly into ranked outcomes.

Cited basis

[5]Pal and Yasar (2024): Multi-agent reinforcement learning is validated for supply-chain disruption and transshipment simulation.

Slide 06

Step 3 - The Output

The system tells you exactly what breaks first.

Breakpoint analysis turns the scenario run into ranked, decision-ready fragility: not just that Port Aurelia is stressed, but which downstream node is most likely to fail first and why.

Demo action

Click "Midwest Logistics Inc." to drill down.

Likely breakpoints

Midwest Logistics Inc.71

FastChip Importers58

PortSide Warehousing42

Regional Bank of Aurelia31

Rank	Node	Type	Breakpoint Probability	Trigger Condition
1	Midwest Logistics Inc.	Tier-2 Trucking	71%	Cash buffer < 7 days (T+14)
2	FastChip Importers	Importer	58%	Margin compression > 15%
3	PortSide Warehousing	Warehouse	42%	Volume drop > 40%
4	Regional Bank of Aurelia	Regional Bank	31%	Loan default cascade

Key insight

The most fragile node is not the port itself. It is a small trucking firm two steps removed, largely invisible to traditional risk monitors.

Cited basis

[11]World Bank Logistics Performance Index (2024): Port dwell-time gaps and indirect logistics dependencies hide system fragility.

Slide 07

Step 3b - Node Deep Dive

Why this node breaks.

The node drill-down makes the causal chain legible: operational dependency, balance-sheet weakness, and behavioral policy all line up to make Midwest Logistics the first absorber of systemic damage.

Node profile

Midwest Logistics Inc.

Tier-2 Trucking / Last-Mile Logistics

System read

This is the first absorber of systemic damage.

Dependencies

Input: 100% of volume from Port Aurelia via PortSide Warehousing.

Output: Serves 3 major electronics manufacturers in the Midwest.

Financials

Cash buffer: 14 days of operating expenses.

Credit line: $2M, fully drawn.

No alternative routing contracts.

Agent policy

If volume drops > 50% for > 7 days, default on equipment leases.

Simulation projection

T+0: Port closes.

T+7: PortSide Warehouse runs out of inventory to ship.

T+14: Midwest Logistics records zero revenue for 7 consecutive days. Cash buffer exhausted. Default triggered.

Slide 08

Step 4 - Burden Transfer Map

Watch the pain move through the system.

PEI visualizes the cascade itself: who absorbs pain first, who passes it on, and which beneficiaries emerge as the system reroutes around the chokepoint.

Demo action

Hover over "AeroCargo Airlines" to see projected +8% stock impact.

Actor map

Burden transfer

T+0

T+7

T+14

T+45

Port Aurelia

Insurers

Midwest Logistics

Manufacturers

AeroCargo Airlines

Graph legend

Red Nodes

Loss absorbers with negative impact

Green Nodes

Beneficiaries with positive impact

Red Edges

Stress transmission such as cost pass-through or volume drop

Green Edges

Value transfer through new demand or margin capture

Edge Thickness

Magnitude of projected impact

Animated time-lapse

T=0: Port Aurelia turns red. Stress pulses outward to ships and insurers.

T=7: Red stress reaches PortSide Warehouse and importers.

T=14: Midwest Logistics flashes bright red and stress reaches Midwest manufacturers.

T=21: Green edges appear as AeroCargo Airlines and alternative ports benefit from rerouting.

T=45: Consumer nodes begin to turn red as price hikes finally pass through.

Cited basis

[6]Anand and Prasad (2024): Supply shocks propagate by reducing output, tightening conditions, and raising inflation.

Slide 09

Step 5 - Intervention Comparator

Test your move before you make it.

The value is not just diagnosis. PEI compares intervention paths, quantifies tradeoffs, and recommends the move that most efficiently restores stability.

Demo action

Click "Apply Intervention C" -> System re-runs the simulation with intervention active.

Intervention	Cost	Midwest Logistics Survival Prob.	Systemic Fragility Index (T+90)	Time to stabilization
A. Do Nothing (Baseline)	$0	29%	94/100	> 90 days
B. Subsidize Air Freight	$400M	52%	78/100	~75 days
C. Targeted Bridge Loans (Logistics Sector)	$200M	88%	52/100	~45 days

Recommended move

"Targeted bridge financing for Tier-2 logistics firms in the Port Aurelia catchment area. Prevents cascade failure. ROI: Positive through tax-revenue stability and avoided economic loss."

Cited basis

[7]Izquierdo et al. (World Bank, 2022): Public investment multipliers are approximately 1.5x in many scenarios.

[8]GFDRR (2019): Resilient infrastructure often returns roughly $4 for every $1 invested.

Slide 10

Step 6 - Post-intervention stabilization path

The new trajectory to stability.

With the selected intervention turned on, the system shows the new stabilization path side by side with the baseline path, including the economic damage that was avoided.

Do nothing path

Midwest Logistics fails at T+14.

Cascading defaults ripple into Midwest manufacturing.

Consumer price hikes reach 12%.

Systemic fragility remains elevated for 90+ days.

Bridge-loan path

Midwest Logistics survives with 88% probability.

Manufacturing continues with minor delays.

Consumer price hike is limited to 4%.

Systemic fragility returns to baseline (52/100) by T+45.

Most probable stabilization path

Liquidity supportLogistics continuityManufacturing output maintainedConsumer inflation moderatedTax base preserved

Cost of instability avoided

$1.2B in lost economic output and social safety-net spending avoided.

Slide 11

Innovation Vector Demo

PEI also models what makes systems stronger.

The same scenario engine can shift from crisis response to strategic investment, showing which innovations permanently lower fragility and reduce future bailout needs.

Demo action

Toggle between "Crisis Response" and "Strategic Investment" to show dual-mode capability.

Government proposes a $500M Digital Green Corridor at Port Aurelia.

Digital customs clearance shrinks from 48 hours to 2 hours.

Participating ships receive subsidized low-emission fuel support.

Metric	Baseline (No Innovation)	With Green Corridor
Midwest Logistics Survival (during stress)	29% -> 88% (with bailout)	94% (without bailout)
Importer Inventory Buffer	14 days required	9 days required (working capital freed)
Systemic Fragility Index	78/100	52/100 (permanent reduction)
Government Fiscal Impact	$200M bailout cost	+$45M tax revenue (increased trade)

Key insight

The innovation pays for itself by reducing the need for future bailouts and raising baseline economic activity.

Cited basis

[9]UN ESCAP and ADB (2024): Digital trade facilitation can reduce trade costs by up to 11%.

[10]Global Infrastructure Hub (2025): A large infrastructure gap makes ROI-led investment prioritization critical.

Slide 12

Backtesting and Calibration

PEI learns from every real-world event.

Backtests turn historical crises into model calibration loops, improving the mechanism library and increasing confidence in first-absorber and beneficiary prediction.

Example replay

2024 Red Sea Crisis

Houthi attacks force rerouting around the Cape of Good Hope, creating a natural validation case for shipping costs, traffic loss, first absorbers, and beneficiaries.

Metric	PEI Predicted	Actual Observed	Accuracy
Shipping Cost Increase	+135%	+141%	96%
Suez Canal Traffic Drop	-72%	-75%	96%
First Absorber (Sector)	European Importers	European Importers	Correct
Beneficiary (Sector)	Air Freight / Southern African Ports	Air Freight / Southern African Ports	Correct

The flywheel effect

Each real-world event improves calibration of the PEI mechanism library.

Accuracy of first-absorber prediction rises as more crises are replayed and scored.

The calibration loop compounds into a proprietary data advantage that is difficult to replicate.

Cited basis

[4]World Bank (2025): The Red Sea crisis drove shipping costs up 141% and cut Suez traffic by 75%.

Slide 13

Technical Integration

Designed for enterprise integration.

PEI is intended to slot into an existing data and risk stack: we ingest proprietary exposures, run simulation workloads, expose outputs through APIs, and support controlled deployment models.

Your Data Sources

AIS shipping data

Market feeds

Internal risk models

PEI Ingestion Layer

Prefect flows

Data validation

Graph population

PEI Core Simulation

Ray cluster

Mesa / LangGraph

Neo4j / Redis

PEI API

REST API

Webhooks

CSV / JSON export

Your Systems

Dashboards

Risk workbenches

Decision workflows

Integration options

API-first: expose scenario runs, ranked breakpoints, and intervention outputs through REST.

Data ingestion: import proprietary exposure data to customize the world-state model.

Co-development: build custom agent policies and transfer mechanisms specific to the client's domain.

On-premise / VPC deployment: support highly sensitive data environments.

Security and compliance

SOC 2 Type II (targeted)

GDPR-compliant handling patterns

Role-based access control

Full audit log of simulation runs and agent decisions

Slide 14

What You Have Seen Today

PEI transforms how teams navigate uncertainty.

The shift is from monitoring to mechanism, from alerts to ranked failure paths, and from reactive commentary to intervention design.

Traditional Approach	PEI Approach
There is a risk in Port Aurelia.	Midwest Logistics Inc. has a 71% probability of failure at T+14.
We should monitor the situation.	A $200M targeted bridge loan reduces systemic fragility by 42 points.
We will react when something breaks.	We can preempt the cascade before the first node fails.
Innovation ROI is hard to quantify.	The Green Corridor pays for itself in 3 years through reduced bailout exposure.

Core value proposition

PEI gives decision-ready intelligence about how systems react under pressure, not just that pressure exists.

Slide 15

Next Steps

Let us build this for your specific domain.

The design-partner motion is scoped around discovery, custom model build, pilot deployment, and ROI evaluation so the product is shaped around the partner's actual exposure network.

Phase	Duration	Deliverable
Phase 1: Discovery	2 Weeks	Map your specific exposure network and identify 3-5 key scenarios.
Phase 2: Model Build	6 Weeks	Build the custom world-state model and calibrate agent policies with your data.
Phase 3: Pilot	8 Weeks	Deploy PEI in your environment and run live scenarios with your team.
Phase 4: Evaluation	2 Weeks	Measure ROI and define commercial terms for full deployment.

Immediate next steps

Schedule a technical deep dive with engineering to review architecture and security.

Share a data-inventory template for mapping the partner's relevant data sources.

Prioritize the top 3 scenarios that matter most in the partner's operating context.

Slide 16

Q&A / Appendix

Thank you.

The appendix makes it easy to continue into technical review, governance review, or deeper design-partner discussion without rebuilding the narrative from scratch.

Name

[Name]

Title

[Title]

[Email]

Phone

[Phone]

Appendix slides available on request

Full technical architecture diagram

Detailed agent policy schema

Security and compliance documentation

Case study: Red Sea Crisis backtest (full report)

Complete citation list (World Bank, IMF, Lloyd's, and more)

References

Distribution-ready source list

Continue the conversation

1.World Bank. (2025). Global Economic Prospects, January 2025.
2.Lloyd's of London. (2024). Systemic Risk Scenario: Geopolitical Conflict and Global Trade Disruption.
3.UNCTAD. (2024). Review of Maritime Transport 2024.
4.World Bank. (2025). The Red Sea Shipping Crisis: Trade and Inflation Impacts.
5.Pal, K., and Yasar, M. (2024). Multi-Agent Reinforcement Learning for Transshipment. IEEE Transactions on Engineering Management.
6.Anand, R., and Prasad, A. (2024). Supply Chain Disruptions and Financial Conditions. Economics Letters.
7.Izquierdo, A., et al. (2022). Is Public Infrastructure Investment Productive? World Bank Policy Research Working Paper.
8.GFDRR / World Bank. (2019). Lifelines: The Resilient Infrastructure Opportunity.
9.UN ESCAP and ADB. (2024). Asia-Pacific Trade Facilitation Report 2024.
10.Global Infrastructure Hub. (2025). Global Infrastructure Outlook.
11.World Bank. (2024). Logistics Performance Index 2023/2024.
12.Bank for International Settlements. (2024). Stress Testing: Principles.
13.IMF. (2024). Geoeconomic Fragmentation and the Future of Trade.
14.Freund, C., et al. (2024). Improving Supply Chain Resilience. World Bank.
15.World Bank Group. (2017-2025). Maximizing Finance for Development Approach.