Risk Radar · GoldShore
360° risk intelligence.
Continuous. Institutional.
Risk Radar aggregates position exposure, political volatility, financial uncertainty, and macro-climate signals into a single, continuously updated risk picture — built for operators who cannot afford blind spots.
Social-ready visual
GoldShore Risk Radar
Live signal sweep across 96 volatility bands, anomaly queues, and directional bias layers — refreshed every 2.4 seconds.
Architecture
Four engines. One risk picture.
Risk Radar is a layered engine stack. Each module scores a different domain of risk independently, and the composite snapshot gives operators a unified threat surface they can act on — from individual position liquidity to macro climate-finance coupling.
Core Engine
Multi-Asset Exposure Engine
Aggregates position data across sports, TCG, and equity asset classes. Computes total exposure, liquidity scores via exponential decay, and blends external threat signals into a composite 360° snapshot updated on every cycle.
Pol-Quant
Political Quantification Engine
Classifies political text — legislative filings, social media, and polling summaries — using ProsusAI/FinBERT. Emits normalised Volatility Scores and Policy Shift Alerts when political instability crosses the defined threshold.
FRA
Financial Risk Assessment
Scores five independent dimensions of financial uncertainty — interest-rate duration, credit default probability, FX exposure, bid-ask liquidity risk, and HHI sector concentration — then blends them into a composite classification with per-dimension rationale strings.
EPO
Economic Psychological Oscillator
Connects NOAA Oceanic Niño Index climate data (El Niño / La Niña phase detection) with VIX, commodity price pressure, consumer confidence, and optional FinBERT economic narrative sentiment into a single −100 to +100 oscillator that reveals when climate cycles and financial psychology are reinforcing or divergent.
FRA · Financial Risk Assessment
Five dimensions of financial uncertainty, scored simultaneously.
Each dimension produces an independent 0–100 score with a machine-readable rationale string and a contributing-factor breakdown. The weighted composite maps to one of four classifications your downstream systems can act on directly.
Interest Rate
25% composite weightWeighted-average modified duration across all bond holdings. A 10-year portfolio duration maps to a score of approximately 80 out of 100.
Credit
25% composite weightS&P-style rating → default risk map (AAA = 2, D = 100), weighted by mark-to-market value. Investment-grade assumed when rating is absent.
Currency
15% composite weightNon-USD portfolio share. Every currency position is identified by ISO 4217 code; 83%+ foreign denomination saturates the score at 100.
Liquidity
20% composite weightWeighted-average bid-ask spread across all positions. A 5% average spread maps to a score of 100. Asset-class defaults are applied when live spread data is absent.
Concentration
15% composite weightHerfindahl-Hirschman Index computed on sector weights. Perfectly diversified converges to near zero; single-sector books score 100.
EPO · Economic Psychological Oscillator
Where El Niño meets the VIX.
EPO is the only module in the market that systematically connects NOAA Oceanic Niño Index data with VIX, commodity price pressure, and consumer confidence — detecting when climate cycles and financial psychology are reinforcing each other or pulling in opposite directions.
NOAA ONI SST anomaly → El Niño / La Niña phase detection
Normalised VIX inverted: 0 VIX → +100, 100 VIX → −100
Food & energy price pressure index (0–100)
Consumer confidence index (0–100), centred at 50
ProsusAI/FinBERT via HuggingFace Inference API — enriches the signal when provided
Who it's for
Built for teams operating at the edge of market complexity.
Portfolio Operators
Continuous 360° risk snapshots across multi-asset books. Composite threat levels update in real time as market volatility and political instability shift — so operators always know the current posture without manual reconciliation.
- Live exposure breakdown by asset class
- Liquidity decay scoring across instruments
- Real-time external threat monitoring
- LOW / ELEVATED / HIGH / CRITICAL labels
Risk Managers
Five-dimensional Financial Risk Assessment gives structured, auditable visibility into interest rate, credit, currency, liquidity, and concentration risk in a single API call — each dimension accompanied by a rationale string and contributing-factor map.
- BENIGN / MODERATE / STRESSED / ACUTE classification
- Per-dimension rationale strings
- Weighted HHI sector concentration tracking
- Audit-ready signal metadata on every result
Quant Developers
Consume FRA results and EPO readings directly from the engine-durable TypeScript package. Pure functions, no runtime side effects, and both async and sync variants mean Risk Radar composites into any downstream signal bus without friction.
- TypeScript-native, fully tree-shakeable
- Async computeEPO + sync computeEPOSync
- FinBERT enrichment is strictly optional
- Published as @goldshore/engine-durable
Access tiers
Start evaluating. Scale when you're ready.
All tiers share the same engine infrastructure. Capabilities unlock as your operational requirements grow — there is no rebuild required when moving between tiers.
Observer
Static Risk Radar snapshot via the public API. Read-only, rate-limited access ideal for evaluation and prototyping.
- Single-account snapshot per request
- External threat level + liquidity score
- Risk level classification label
- Public REST endpoint
Most popular
Operator
Full Risk Radar engine with FRA, Pol-Quant, and real-time signal delivery — built for production deployments.
- Everything in Observer
- FRA — all 5 dimensions + composite score
- Pol-Quant political volatility scoring
- Webhook signal delivery < 28ms
- Multi-account aggregation
- Audit log with full signal metadata
Institutional
EPO climate-finance coupling, custom model weighting, white-label deployment, and an SLA-backed support relationship.
- Everything in Operator
- EPO — Economic Psychological Oscillator
- Custom ONI + VIX + commodity feed integration
- White-label API and dashboard
- Dedicated SLA + support engineer
- On-prem or private cloud deployment
Common questions
Answers before the briefing.
- How does Risk Radar connect to my existing data?
- Risk Radar ingests position data, market volatility readings, and political text through a structured API. The engine-durable package accepts typed TypeScript inputs so integration is straightforward from any data pipeline.
- What does the EPO actually predict?
- EPO does not make predictions — it produces an oscillator reading that reflects the current alignment between macro-climate signals (NOAA ONI) and financial-market psychology (VIX, commodity pressure, consumer confidence). A divergent reading flags when these domains are moving in opposite directions, which historically precedes regime shifts.
- Is FinBERT scoring required for the FRA or EPO?
- No. FinBERT scoring is an optional enrichment layer. The FRA runs entirely on quantitative position inputs. The EPO runs on ONI + market data; FinBERT economic-narrative sentiment is additive when an HuggingFace token is supplied.
- Can Risk Radar operate on non-financial assets?
- Yes. The core engine natively supports sports cards, trading card game (TCG) collections, and equity positions alongside standard financial instruments. Asset classes are typed and extensible.
- What is the sub-28ms latency figure based on?
- The < 28ms figure reflects p95 signal pipeline latency from position data ingestion to composite snapshot delivery, measured on the Cloudflare Workers edge infrastructure that powers the GoldShore API.
Next step
Get a Risk Radar briefing.
We scope every engagement to your actual data environment, latency requirements, and downstream decision systems. No generic demos — only what applies to your operation.