Some false stories die on their own. Others move markets and mobilize people. We'll tell you which one is targeting your operations.
Real-time economic nowcasting meets network analysis, so you know when a narrative is manufactured, whether it contradicts the facts, and whether it will stay online or spill into the real world.
The problem
You can't tell coordinated from organic by looking at individual posts. Without graph-level analysis, manufactured amplification looks like genuine public concern.
Narratives claim prices are spiking, supply is failing, or policy has shifted - but is that actually happening? Without real-time economic data as a baseline, you can't measure the gap between what's said and what's real.
Most online narratives stay online. Some trigger protests, market moves, or policy pressure. Without a model of real-world impact likelihood, teams either over-react or miss the moment to act.
How it works
1.Coordinated or organic?
Graph analysis classifies whether a narrative is spreading through genuine public interest or manufactured coordination - synchronized posting patterns, shared infrastructure, and amplification networks reveal the difference.
2.Nowcast economic reality
Real-time economic nowcasting builds a continuously updated picture of what is actually happening - prices, supply, demand, policy indicators - so the system has ground truth to compare narratives against.
3.Measure the narrative–reality gap
Each narrative is scored against nowcast data to quantify how far the claims diverge from measurable reality. A large gap in a coordinated campaign is a strong signal; a small gap in organic discussion is not.
4.Assess likelihood of real-world impact
Network-diffusion modeling estimates whether the narrative is likely to stay online or tip into collective action - protests, market moves, policy pressure - and how soon.
5.Deliver a decision recommendation
All signals converge into a single recommendation - react, monitor, or escalate - with explicit confidence levels, geographic risk mapping, and a full audit trail.
What makes us different
Economic nowcasting as ground truth
We maintain a real-time model of what is actually happening in the economy - prices, supply chains, macro indicators. This gives us an objective baseline that no amount of narrative manipulation can shift, and lets us quantify exactly how far a story diverges from reality.
Coordinated vs. organic classification
We analyze the network structure behind how a narrative spreads. Coordinated campaigns leave graph-level fingerprints - synchronized timing, shared amplification infrastructure, cluster topology - that are invisible at the level of individual posts.
Real-world impact likelihood
We model whether a narrative will stay online or spill into collective action that impacts people - protests, market runs, policy shifts. The model estimates how likely, how soon, and where - so you act at the right moment, not too early and not too late.
Research pedigree
Our detection and action-risk models are built on published research in network science and economics, not heuristics.
Read our whitepapersTeam affiliations & background
Polish Academy of Sciences
(PAN)
Network-science and game-theoretic research
SGH Warsaw School of Economics
Quantitative economics and nowcasting methodology
International Monetary Fund
(IMF)
Macroeconomic and institutional experience
National Science Centre
(NCN - Narodowe Centrum Nauki)
Nationally funded research grants
Who it's for
Energy & critical infrastructure
Purpose-built for risk and communications teams in energy - where narratives about pricing, supply security, and policy directly affect operations.
Governments & public institutions
Situational awareness for government analysts monitoring influence operations that target public trust, elections, or national cohesion.
Cyber forces & defense
Operational intelligence for cyber commands and defense units countering information warfare as part of the hybrid threat landscape.
Sector-agnostic architecture
The detection, nowcasting, and diffusion-modeling stack is domain-independent. Energy is our first vertical; the methodology applies wherever coordinated narratives meet measurable economic reality.
Trust & responsibility
Human in the loop
Every output is a recommendation for a human decision-maker. The system never takes autonomous action.
Explicit confidence levels
All assessments carry calibrated confidence scores. You always know how certain the system is - and where it is uncertain.
Full audit trail
Every detection, score, and recommendation is traceable to its inputs. Reproducible, explainable, reviewable.
On-prem & data sovereignty
Deployable within your infrastructure. Data never leaves your jurisdiction without explicit authorization.
EU AI Act aligned
Designed from the outset for compliance with the EU AI Act's requirements for transparency, human oversight, and documentation.
Decision support, not automation
We provide intelligence and calibrated recommendations. We do not make decisions. That distinction is architectural, not aspirational.