Data Quality and Freshness

Freshness and quality are different properties.

• Freshness asks whether the data is recent enough for the intended use.
• Quality asks whether the data is complete, consistent, valid, and suitable for processing.

A recent candle series can be incomplete. A complete filing can be several days old but still relevant historical evidence.

Timestamps

NataPulse may expose:

• source occurrence or publication time;
• ingestion time;
• model generation time;
• last update time;
• report generation time.

Users should read the timestamp that corresponds to the evidence, not only the time the page was refreshed.

Quality states

Structured market and quantitative data can use states such as:

• Ready: sufficient for the intended calculation.
• Degraded: usable with a known quality issue.
• Stale: last data is older than the allowed window.
• Incomplete: required observations are missing.
• Invalid: the data failed validation.

Source-dependent cadence

There is no single NataPulse refresh interval. Examples:

• live event channels can update as events are published;
• a page may fall back to periodic polling;
• market datasets can be daily, delayed, or intraday;
• whale summaries may use a separate polling cadence;
• SEC and news latency depends on the upstream source;
• quantitative generation follows its configured windows.

Honest failure states

Product pages should show one of:

• current data;
• a visible delayed or degraded status;
• an empty state;
• an error state;
• the last known data with its timestamp.

They should not generate synthetic activity to make a page appear populated.

For practical interpretation, see Data Freshness Reference.