Most AI retrieval systems surface what you already know to look for. Merlin Intelligence is built for a different problem.
We develop an open-source graph intelligence platform that applies ontology learning, spectral graph theory and convex optimisation to heterogeneous data — documents, structured feeds, and time series — to identify three classes of insight that standard pipelines systematically miss:
· Thematic Representatives — the dominant narrative poles structuring a dataset
· Hinge Connectors — the conceptual bridges between thematic clusters
· Frontier Themes — the weak signals and non-consensus indicators that nothing else echoes
The platform is designed for investment banks, institutional asset managers, macro strategists, climate finance practitioners, and strategic advisory teams operating at the intersection of geopolitics, markets, and complex negotiation.
It runs on locally retrieved subgraphs, preserving data sovereignty and keeping the analyst in control. Every output is a named, interpretable signal — not a black box, not a chatbot response.
We work with decision makers who need not just retrieval but orientation: a structured map of what matters, what bridges the debate, and what the data is not yet saying.
Intelligence between signal and decision.

Merlin Intelligence

Org chart

Teams

Offices

Merlin Intelligence

Org chart

Teams

Offices

Related companies