Loading
NEXUS is a geopolitical-market convergence intelligence platform. It detects alpha signals across layers that institutional desks currently track in isolation — and delivers them as backtested, executable theses.
A macro desk watches Fed language. A geopolitical team tracks Middle East escalation. A quant monitors VIX structure and calendar anomalies. These teams rarely sit in the same room, let alone feed their signals into a shared model.
The result: positions sized on incomplete pictures. Events that were visible in advance — in the convergence of signals — arrive as surprises. Capital is deployed reactively instead of ahead of the move.
Geopolitical, macro, and OSINT signals tracked by separate teams with no cross-synthesis
Market positioning decisions made without access to real-time OSINT or escalation modeling
No systematic way to score, backtest, or validate analyst predictions against market outcomes
Calendar and event-driven context largely siloed from quantitative signal analysis
NEXUS runs a continuous convergence engine across geopolitical, market structure, OSINT, and systemic risk layers, with calendar and celestial data as narrative/actor-belief context. When primary signals align, the system scores the event, generates a structured thesis, and tracks it through to outcome.
Live military aircraft tracking, GDELT OSINT feeds, conflict zone mapping. Escalation patterns visible before they reach the wire.
Convergence events scored 1–5 across 4 primary layers + narrative overlay. Intensity threshold alerts. Full signal history with market correlation.
Claude-powered analyst with 20+ live tools: market regime, options flow, on-chain, shipping intelligence, game theory modeling.
Time-gated prediction simulation. Brier scoring, log-loss, p-value significance testing against random baseline. No look-ahead bias.
Manual portfolio tracking live today. Direct broker integrations (Trading 212, Coinbase, and more) coming soon.
pgvector-powered semantic memory. Actor profiles, world models, historical theses — searchable and referenced in every AI response.
| Dimension | NEXUS | Incumbents (Bloomberg, Reuters Eikon) |
|---|---|---|
| Signal synthesis | 4 primary layers + narrative overlay fused into scored convergence events | Data terminals; no cross-layer synthesis |
| Prediction validation | Brier score, p-value, time-gated backtests | Analyst reports; no systematic accuracy tracking |
| Geopolitical + market | Unified: war room feeds directly into thesis | Separate products; no causal linkage |
| Calendar signals | Hebrew, Islamic, FOMC, OPEX, Chinese convergence | FOMC dates only; no esoteric calendar research |
| Execution integration | Direct broker execution (coming soon) + manual portfolio tracking | Data only; no execution layer |
| Cost | $150–$450/mo (institutional: custom) | $2,000–$25,000+/mo per seat |
The global financial data and analytics market is valued at over $40B, growing at 12% annually. Bloomberg and Refinitiv capture the majority of institutional spend — at price points that make them inaccessible to the long tail of professional allocators, family offices, and independent funds.
The alternative data segment — geopolitical risk, OSINT feeds, satellite imagery, sentiment analytics — is separately valued at $2.7B and growing at 28%. NEXUS sits at the intersection, combining these data types into a single synthesis layer at a fraction of incumbent cost.
Near-term revenue path is individual and small-fund SaaS. The Institution tier and API licensing represent the high-value expansion opportunity — embedding NEXUS signals into existing quantitative workflows at fund level.
The signal convergence engine is proprietary — a multi-layer scoring system that correlates geopolitical, celestial, economic, and OSINT events across historical time series to detect statistically meaningful convergence points.
The AI layer runs on Anthropic Claude, with Voyage AI embeddings and pgvector for semantic retrieval. The architecture is built for institutional-grade reliability: Next.js 15, PostgreSQL on Neon (serverless), Stripe for billing, and a modular API layer that supports white-label and custom data integration.
Senior Software Architect with 20 years of experience across financial services, data infrastructure, and platform engineering. Founded Polyxmedia as a product studio building intelligent systems at the intersection of data and capital markets.
NEXUS is the product of a conviction that geopolitical signal synthesis can be done systematically, and that the gap between intelligence and execution is a solvable engineering problem.
London-based product studio. Specialises in intelligent systems, data platforms, and financial technology. NEXUS is the flagship product.
If the opportunity makes sense, I would rather spend 30 minutes in a live demo than send a deck. You can see exactly what the platform does, what the data looks like, and where it is going.