Where innovation meets collaboration.

Typical scenario

A sovereign macro strategist needs to know whether Brazil's debt path stays financeable if growth disappoints and the Selic has to stay high.

Example prompt

“Simulate Brazil 2026-2050 with GDP growth 1pp below consensus and Selic held at 12% for three years. Show the debt/GDP distribution and the probability of a fiscal limit breach.”

What you get back

• —Fan chart of debt/GDP across 500 Monte Carlo paths with 5/50/95 percentiles
• —Probability of breaching the fiscal limit before 2035, by scenario
• —Household stress and FX pass-through decomposition per year

Typical scenario

A risk committee wants a same-day read on how an oil shock would travel through their country exposures.

Example prompt

“Oil jumps to $120 and stays there for two quarters. Which countries in my book deteriorate first, and how far does it spread?”

What you get back

• —Ranked country risk deltas with the transmission channel named for each
• —Contagion map showing first-, second- and third-order spillovers
• —Scenario narrative plus the assumptions the model actually used

Typical scenario

An operations lead has to decide where to add buffer capacity before the next outage season.

Example prompt

“Take my pipeline and refinery network, fail the two highest-throughput nodes independently, and rank mitigations by served demand recovered per dollar.”

What you get back

• —Criticality ranking of nodes and links by unserved demand
• —Before/after flow maps for each mitigation option
• —Cost-effectiveness table with recovery time under each strategy

Typical scenario

A commodities trader is sizing a US Gulf to Asia LPG position and wants the arb window and downside quantified.

Example prompt

“With freight up 20% and Asian propane at a $90/t premium, is the US Gulf to Chiba arb open? Size a position and show its VaR.”

What you get back

• —Arb window open/closed by week with the netback build-up
• —VLGC fleet utilisation and voyage economics per route
• —Position P&L distribution with 95% and 99% VaR

Typical scenario

A logistics planner needs to know whether the current PSV fleet can keep FPSO diesel inventories safe through a rough-weather month.

Example prompt

“Run the Santos Basin cluster for 90 days with 30% more weather downtime. Do any FPSOs go below minimum diesel, and does one more PSV fix it?”

What you get back

• —Inventory trajectories per FPSO with stock-out events flagged
• —PSV utilisation, queueing and waiting-on-weather hours
• —Fleet-size sensitivity showing the marginal value of each added vessel

Typical scenario

A revenue manager wants to know what happens to yields if a competitor undercuts fares on a core route.

Example prompt

“A competitor cuts base fares 15% on my top three routes. Simulate passenger switching and show my optimal fare response.”

What you get back

• —Load factor and yield paths with and without a response
• —Passenger switching flows by fare class and booking curve
• —Recommended fare response with the revenue delta it buys

Typical scenario

A retail bank is choosing between two next-best-action strategies and wants them tested on synthetic households first.

Example prompt

“Compare a savings-first versus a credit-first next-best-action policy across UK household avatars under a mild recession.”

What you get back

• —Uplift by household segment for each policy
• —Arrears, attrition and balance-growth curves under the macro scenario
• —Policy comparison with the segments where each one wins

Typical scenario

An allocator wants to know which hedge sleeve actually protects purchasing power if the monetary regime shifts.

Example prompt

“Stress USD purchasing power under fiscal dominance, and score gold, TIPS, foreign equity and crypto sleeves over ten years.”

What you get back

• —Purchasing-power paths per monetary regime with tail percentiles
• —Hedge sleeve scorecard: real return, drawdown, regime hit rate
• —Contagion trace from Treasury issuance to household balance sheets

Typical scenario

A quant wants to see how a mean-reversion strategy behaves in a market of reacting traders, not a replayed tape.

Example prompt

“Calibrate to last year's volatility and run my mean-reversion strategy against momentum and market-maker agents. Where does it break?”

What you get back

• —Calibration diagnostics against observed volatility and volume
• —Strategy P&L distribution with drawdown and turnover
• —Regime breakdown showing the conditions where the edge disappears