6 unique trading signals from 60+ sources. Entity-level sentiment, source divergence, and contrarian detection that Bloomberg and RavenPack don't offer — at 1/10th the price.
Each signal is designed for a specific edge that institutional traders actually pay for.
Per-entity scores tracked across 60+ sources — not just article-level polarity. Know exactly how NVIDIA, the Fed, or Bitcoin sentiment is shifting.
When Reuters says bullish but CNBC says bearish — that disagreement IS the alpha. We surface it automatically with severity scoring.
Credible minority opinions bucking consensus. When one high-quality source diverges from the pack, you see it before the market does.
Co-occurring entities forming emerging themes. Detect 'AI regulation wave' or 'crypto contagion' before it's priced into the market.
Rate of change + acceleration of entity sentiment. Momentum matters more than absolute level for timing entries and exits.
Reuters (0.95) matters more than Reddit (0.55). Every score is weighted by source quality — most tools treat all sources equally.
Features that institutional buyers actually evaluate — and where we lead.
| Capability | RavenPack $50k+/yr | Bloomberg $24k/yr | Refinitiv $22k+/yr | Sentifi $10k+/yr | MktPeek $5.9k/yr |
|---|---|---|---|---|---|
| Entity-level sentiment | ✓ | ◐ | ✓ | ✓ | ✓ |
| Source divergence detection | ✗ | ✗ | ✗ | ✗ | ✓ |
| Contrarian alerts | ✗ | ✗ | ✗ | ✗ | ✓ |
| Narrative clustering | ✗ | ✗ | ✗ | ◐ | ✓ |
| Sentiment velocity | ✗ | ✗ | ✗ | ✗ | ✓ |
| Credibility weighting | ✓ | ◐ | ◐ | ✗ | ✓ |
| Multi-engine ensemble (3+) | ✓ | ✓ | ✓ | ✗ | ✓ |
| FinBERT / Transformer NLP | ✓ | ✓ | ✓ | ✗ | ✓ |
| Crypto + Macro + Equities | ◐ | ✓ | ◐ | ◐ | ✓ |
| Prediction market correlation | ✗ | ✗ | ✗ | ✗ | ✓ |
| SEC filing timing signals | ✓ | ✓ | ✓ | ✗ | ✓ |
| REST API access | ✓ | ✓ | ✓ | ✓ | ✓ |
| Real-time streaming | ✓ | ✓ | ✓ | ✗ | ◎ |
| Historical data export | ✓ | ✓ | ✓ | ◐ | ◎ |
RSS feeds, financial APIs, SEC filings, prediction markets — continuously collected via 24 automated jobs.
Every article scored by keyword analysis, pattern matching, AND FinBERT (financial BERT transformer). Ensemble weighted by confidence × source credibility.
70+ tracked entities (companies, crypto, macro indicators) extracted and scored individually across all sources.
Source divergence, contrarian alerts, narrative clusters, and sentiment velocity computed in real-time.
Clean REST API with 11 endpoints. JSON responses, < 500ms latency. Plug into your existing quant stack.
11 endpoints. JSON responses. Plug into any quant stack.
GET /v1/entities # All tracked entities with sentiment scores
GET /v1/entities/{entity} # Deep dive: timeline, sources, alerts
GET /v1/signals/divergence # Active source disagreements (alpha)
GET /v1/signals/contrarian # Credible minority opinions
GET /v1/signals/velocity # Sentiment momentum leaders/laggers
GET /v1/signals/clusters # Emerging narrative clusters
GET /v1/briefs/latest # Auto-generated market intelligence brief
GET /v1/markets/kalshi # Prediction market snapshots
GET /v1/markets/polymarket # Decentralized prediction markets
GET /v1/articles # Scored article feed (paginated)
GET /v1/stats # Pipeline health & coverage stats{
"entity": "Meta",
"divergence": 1.1314,
"most_bullish": { "source": "Financial Times", "score": 0.65 },
"most_bearish": { "source": "Stephen Council", "score": -0.48 },
"source_count": 8,
"signal": "STRONG_DIVERGENCE"
}RavenPack charges $50,000+/yr. Bloomberg Terminal is $24,000/yr.
We deliver more unique signals at a fraction of the cost.
For individual researchers & small funds
For quant teams & mid-tier funds
For institutional desks & research teams
Source divergence, contrarian alerts, and narrative clustering — capabilities that RavenPack, Bloomberg, and Refinitiv simply don't offer at any price.