Multi-Source Intelligence

Prism doesn't rely on a single data feed. The AI simultaneously processes five categories of market intelligence.

Data Sources

Price Data

Candlestick data

Real-time OHLCV across multiple timeframes

Volume analysis

Trading activity and momentum indicators

Order book depth

Buy/sell pressure and support/resistance levels

On-Chain Data

Whale movements

Large wallet transfers and accumulation patterns

Token transfers

Unusual activity and distribution changes

Contract interactions

Smart contract calls, approvals, and liquidity events

LP Metrics

Pool depth

Available liquidity for trading

Impermanent loss

LP position health indicators

TVL changes

Protocol confidence and capital flow direction

News & Social

Crypto news sentiment

Positive/negative/neutral scoring of headlines

Twitter/X trends

Community discussion topics and sentiment shifts

Community signals

Telegram/Discord activity and engagement patterns

Macro Context

BTC dominance

Risk-on vs risk-off market regime

Total market cap

Overall crypto market health

DeFi TVL shifts

Capital rotation between protocols and chains

Data Fusion

The 70B model doesn't just aggregate signals — it synthesizes them. When price data shows a dip but on-chain data shows whale accumulation and news sentiment is neutral, the AI recognizes this as a potential buying opportunity rather than a simple bearish signal.

CONFLICT RESOLUTION

When data sources conflict, the AI weighs each signal by reliability and recency, then forms a probabilistic view rather than a binary buy/sell decision.