Interactive Demo

Explore how signals are generated and explained

This page simulates a signal review workflow. Use the controls to see how different inputs change the label and the explanation summary. The values shown are illustrative and are intended to teach methodology and terminology.

Demo controls

Educational

Adjust the inputs below. The model on this page uses a simple, explainable rule set to illustrate how a monitoring system can convert observations into a structured signal summary.

Time window

In practice, signals are often different across windows because volatility and noise change.

Trend strength

Weaker 55 Stronger

Volatility level

Lower 42 Higher

Order flow pressure

This represents an example feature derived from microstructure style inputs.

Historical similarity

A similarity score can be used to highlight whether conditions are typical or unusual.

This demo does not collect personal data. Your selections are processed locally in your browser to illustrate how an explainable signal summary can be produced.

Prefer a deeper walkthrough? Visit Resources for definitions and examples.

How to read the output

The output cards summarize three layers: the label (a compact name), the rationale (why it was assigned), and the suggested review actions (how an analyst might verify it). A monitoring platform often separates detection from decision making so the review can stay consistent.

Labels are descriptive categories, not predictions.
Rationales list the strongest contributors to the label.
Review actions emphasize cross checks and context.

Signal workspace (simulated)

A simplified view showing how monitoring inputs map to a structured, explainable summary.

Local calculation

gold market analysis demo chart with trend line and signal markers

Window

Daily

Trend

Volatility

Current label

Balanced watch

Conditions appear mixed. The workflow suggests monitoring for confirmation rather than acting on a single input.

Moderate

Rationale (top contributors)

Trend and volatility are within a mid range, which often reduces confidence in single factor interpretations.
Order flow is balanced, suggesting that directional pressure is not dominant in this view.
Historical similarity indicates a common setup, which is useful for comparison but not decisive by itself.

Suggested review actions

Check nearby time windows (for example, 4h vs daily) to confirm whether the structure persists.
Review the feature inputs for data gaps and consistent timestamps before interpreting the label.
Compare with similar historical periods and record what differed when outcomes diverged.

See feature details Browse use cases

From raw inputs to features

Monitoring systems typically standardize inputs first. Normalization reduces inconsistencies across sources and helps ensure that each feature is comparable across time and market regimes. In a production platform, this step also includes validation to detect gaps, duplicates, and timestamp drift.

In this demo, the sliders represent a small set of features. A real workflow may compute many features, but the principle is the same: define them clearly, keep them stable, and document how they are calculated.

Explainability as a default

A useful signal is not only a label, but also a record of why the label was assigned. Explanations support review, reduce confusion when conditions change, and help teams communicate consistently. When explanations are missing, a signal can be difficult to trust or reproduce.

The output on this page includes rationale and suggested review actions so you can see how a monitoring tool can support a structured decision making process without pushing a specific outcome.

Demo disclaimer

This demo is for informational and educational purposes only. It does not provide financial, legal, or investment advice. The signal labels and thresholds shown are illustrative and are not intended for trading decisions.

For additional site disclosures, review our Terms and Privacy Policy.