Financial Intelligence Powered by Knowledge Graphs
Uncover deep insights with GraphRAG. By unqifying S&P 500 equities, options flow, commodity futures, EIA energy data, SEC filings, government contracts, prediction markets, and CFTC positioning into one unified knowledge graph, we reveal the hidden connections driving markets—from insider trading signals to commodity correlations.
Scroll to explore
Why Knowledge Graphs?
Traditional databases see data in silos. Graphs see connections.
Connected Data
Every company links to market data, options flow, government contracts, SEC filings, commodity futures, EIA energy data, and prediction markets—19 collections, 22 edge types
Semantic Search
Find contracts mentioning "AI" or "cybersecurity" using 1536-dim vector embeddings and cosine similarity—no keywords, pure meaning
Hybrid Intelligence
ArangoDB graph queries + Perplexity web search run in parallel. Historical data meets real-time news in under 3 seconds
Intelligence Terminal
What can I help you discover?
Ask about markets, companies, contracts, commodities, or options flow
Graph Architecture
Everything connects to everything. Click nodes to explore how data flows through the knowledge graph.
💡 Click any node to view collection details, schema, and sample queries
Built on ArangoDB with GPT-4 query planning and Perplexity web search
Data Universe
Explore all database collections and their connections
Document Collections
About KARGA
KARGA combines Knowledge Graphs, Retrieval Augmented Generation, and Semantic Search to provide AI-powered financial intelligence across equities, commodities, options, contracts, filings, and prediction markets.
Ask questions in natural language, and GPT-4 generates precise AQL graph queries across 19 interconnected collections (2M+ documents, 22 edge types)—no hallucinations, only real data. Detect insider trading with daily options flow monitoring, correlate commodity prices with company exposure via direct graph edges, analyze sentiment from 7.5K SEC filings, and track whale positioning across 18K prediction markets.
Multi-Source Data
S&P 500 stocks (2M+ OHLCV records), government contracts (500K+), options flow (612 tickers daily), commodity futures (64K CME prices), EIA energy data (crude/natgas inventory), SEC filings (12 form types), prediction markets (18K+ Polymarket/Kalshi), CFTC positioning, and FRED economic indicators—19 collections connected by 22 edge types in one unified knowledge graph.
AI Query Generation
GPT-4 converts natural language into optimized AQL graph queries with semantic search (cosine similarity on 1536-dim embeddings), multi-hop traversals across 22 edge types (e.g., Company → COMPANY_TRADES_COMMODITY → futures_prices → INVENTORY_AFFECTS_PRICE → eia_crude_inventory), insider trading detection via OPTIONS_BEFORE_AWARD/FILING edges, and parallel web search execution.
Blazing Fast
Query 2M+ OHLCV records, 500K+ government contracts, 64K commodity futures, 7.5K SEC filings, 18K+ prediction markets, and 612 daily options flow records with 50ms graph traversals. ArangoDB parallel execution + Perplexity web search delivers complete answers in under 3 seconds.