A forecast you can act on
Each forecast includes a central estimate and a credible interval, showing the expected outcome and the uncertainty around it.
Put foresight to work
Understanding global events through market implied intelligence.
Ask a question about the future and get a probability you can examine. PolyBridge brings together forward-dated market data and a causal model to show the evidence and drivers behind each answer, and how the outlook changes under different assumptions.
Backed by
20VC
Harry Stebbings, Kieran Hill
KDX
Ashby Monk, Elliott Donnelley
Carles Reina
Charles Clancy
CTO, MITRE Corporation
PolyBridge Features
Ask beyond any single market
PolyBridge combines market data, news and your private sources in an explicit causal model to answer conditional and counterfactual questions.
Probability with uncertainty
Get a calibrated probability with an uncertainty range, not just a single number.
Traceable reasoning
Trace each forecast to its contributing sources, with an explicit reasoning chain and a full audit trail.
Scenarios, live updates and alerts
Change assumptions to see how probabilities move, track updates as markets change, and set thresholds for notifications.
Adjust an input. See the downstream effects
Illustrative scenario. Probabilities, ranges, market weights and chart paths are examples.
Ask
Ask
“If the Fed holds rates at its next meeting, what happens to emerging-market sell-off probability?”
This illustrative scenario shows how a conditional question connects to contributing markets.
62%
The interface
Trace each answer to its source markets
Each probability traces to the markets that informed it, with contributions weighted by liquidity.
31% Credit Stress. 19% Geopolitical Escalation
Each driver is a node in the causal model. Condition on it, intervene, or test a counterfactual.
World events are connected
COVID → Supply Chain Issues
- lockdowns
- supply chains break
- stimulus surge
- inflation returns
- rates reset
Russia Invades Ukraine → Energy Shock
- war shock
- energy spike
- food prices
- European stress
- global inflation
SVB → Credit Suisse → Credit Crunch
- deposit flight
- bank contagion
- policy backstop
- credit crunch
- recession risk
Trump Returns → Tariff Risk
- second term
- tariff threat
- supply chains reprice
- inflation risk
- dollar strength
DeepSeek → Tech Sell-off
- model shock
- GPU demand questioned
- Nvidia drawdown
- AI capex repriced
- power demand reset
Strait of Hormuz → Crude Spike
- chokepoint risk
- shipping premia
- crude spike
- inflation path shifts
- EM stress
S&P 500 → Record highs
- Record highs
- valuation questions
- earnings outlook
- rate sensitivity
- what comes next?
Inflation persists → Fed raises rates
What could this mean for inflation and growth in 2027?
Enquire
Common questions
How PolyBridge fits into forecasting, monitoring, and decision work.
PolyBridge returns streamed probability distributions. Uncertainty is explicit. Causal drivers behind the result are reported. Quantitative answers are traced back to forward-dated market data and other sources.
Yes. PolyBridge answers have mechanistic explainability by design, rather than narrative explainability.
Forward-dated market data covers a growing number of event contracts, but those contracts still address a finite set of questions. PolyBridge connects them to answer questions no single contract covers.
LLMs excel at providing compelling narrative, whereas PolyBridge returns streamed quantitative probabilities conditioned on markets. The use cases are different and complementary.
PolyBridge maintains a continuously updated latent model of world state. It's mechanistically interpretable and supports do-calculus on named concepts.
We are continuously adding new data sources, which improves model calibration with every resolved event.
Every answer is a structured probabilistic output. It contains:
- A calibrated probability with a confidence interval. A range that reflects uncertainty, sized to the strength and freshness of the underlying evidence.
- An evidence trail. Each connection is legible. You can see which upstream events drive the answer and by how much.
- Source markets with contribution weights. Every market that informed the answer is named, with its influence shown explicitly.
- A public-data prior, separate from market signal. What historical reference data implies versus what live markets are pricing. Divergence between the two is shown.
- Support and freshness metadata. How many resolved historical examples back the estimate, and when the underlying data was last updated.
This structure means every answer can be reviewed and defended.
Additional documentation is available to qualified institutions upon request. Contact PolyBridge
Teams making high-stakes decisions under uncertainty across defense and intelligence, long-horizon investing, financial services, energy and commodities, and corporate strategy. They use Situation Rooms to monitor live crises, the Forecast Interface to stress-test decisions, and the API to embed foresight into their own systems.
Request access for Situation Rooms, the Forecast Interface, Search, Forecast, and approved RVOL data endpoints, agent workflows, or other PolyBridge tools.