Zero parameters.
Infinite domains.
A signal processing engine with no tunable parameters. Same algorithm for search relevance, anomaly detection, and auto-scaling. Nothing to configure. Nothing to train.
Send a search query and document. The API returns relevance scores with automatically weighted fields. No boosts to configure.
Search Ranking
Auto-weighted field scoring for any document schema. Replaces manually tuned Elasticsearch boosts. Adapts to schema changes without re-tuning.
Anomaly Detection
Three-channel self-checking identity detects anomalies from the first sample. No baseline period. No calibration data. Works on sensors, queries, or financial signals.
Auto-Scaling
Four-band frequency decomposition of traffic patterns. Distinguishes real surges from bot attacks. Instant scaling decisions with confidence scores.
Signal Analysis
Spectral decomposition into natural modes of any time-series. Returns energy distribution, compression ratio, and reconstructed signal.
Complexity Classification
Automatically classifies signals as smooth, structured, complex, or chaotic. Returns optimal processing depth for downstream systems.
Self-Monitoring
The API monitors its own computational integrity using the same anomaly detection it provides to clients. Parameters fitted: zero.
// Add spectral relevance to your Elasticsearch pipeline const response = await fetch('https://sfiniti.com/relevance', { method: 'POST', headers: { 'Content-Type': 'application/json' }, body: JSON.stringify({ query: userQuery, document: esDocument }) }); const { score, field_scores, complexity } = await response.json(); // Use as script_score in your ES query // No field boosts to maintain. No relevance engineering.