Skip to content

Market Data Pipeline

The market data pipeline provides structured price and volume context for assets supported by the active providers.

Assets are mapped to canonical NataPulse keys and provider-specific symbols. Asset classes can include crypto, equities, indices, foreign exchange, and commodities where coverage exists.

Provider data is normalized into canonical OHLCV candles:

  • timestamp;
  • open;
  • high;
  • low;
  • close;
  • volume or amount when available;
  • provider and asset identity;
  • timeframe.

Stable uniqueness rules make candle ingestion idempotent.

Before a candle window is used for quantitative inference, the pipeline evaluates:

  • expected and actual candle count;
  • missing periods;
  • duplicate periods;
  • last candle time;
  • freshness;
  • overall quality score and status.

Quality states include ready, degraded, stale, incomplete, and invalid.

Market observations can become events when they represent relevant movement, anomaly, or context. They can support:

  • Live Pulse and Event Explorer;
  • Market Desk tiles, history, and anomaly views;
  • cross-source clusters;
  • reports and Deep Research;
  • quantitative-signal confirmation.

Latency depends on the asset, timeframe, provider, market calendar, and account configuration. A page refresh interval does not imply that the underlying source is real-time. The interface displays the available source timestamp.

Provider symbols, corporate actions, market holidays, delayed feeds, missing volume, exchange differences, and illiquid periods can affect interpretation. Users should verify critical price decisions against an appropriate market-data source.