ticktickclock.com ticktickclock.com/docs/index.md
Welcome to the documentation hub for TickTickClock — the temporal intelligence layer for the [&] Protocol ecosystem. TickTickClock provides time-series anomaly detection, foreca…

TickTickClock Documentation

Every tick is a learning opportunity. Every pattern has a rhythm. Temporal intelligence that never forgets.

Part of the [&] Protocol stack · Ecosystem overview · Three-protocol stack · Stack status

Welcome to the documentation hub for TickTickClock — the temporal intelligence layer for the [&] Protocol ecosystem. TickTickClock provides time-series anomaly detection, forecasting, and pattern recognition as continual learning services.

TickTickClock is the &time primitive provider, exposing &time.anomaly, &time.forecast, and &time.pattern capability contracts. It does not store business data, host agents, or manage workflows — it ingests temporal signals, learns their structure continuously, and returns intelligence to upstream consumers.

The Problem

Time-series intelligence in production systems is fragmented. Anomaly detection runs as a batch job. Forecasting lives in a notebook. Pattern detection is a dashboard someone checks on Mondays. None of these systems learn continuously, and none of them talk to each other.

Meanwhile, underlying models suffer from catastrophic forgetting — retrained monthly on a window of recent data, discarding everything about long-range seasonality. Multi-timescale temporal behavior requires multi-timescale memory, and no production system provides it.

TickTickClock solves this with:

  • Continual learning — models update at inference time, no batch retraining

  • Multi-timescale memory — fast (seconds), medium (hours), slow (days), glacial (months)

  • Streaming-first — designed for unbounded streams, not fixed datasets

  • Edge-capable — SQLite + Bumblebee on constrained devices

Documentation Map

Homepages

Root Docs

TickTickClock Docs

[&] Capability Contracts

CapabilityOperationsDescription
&time.anomalydetect, enrich, learnAdaptive anomaly detection via selective state-space models
&time.forecastpredict, explain, enrich, learnMulti-timescale forecasting with continual learning
&time.patterndetect, summarize, enrichCycle and motif detection via spectral analysis

MCP Tools

ToolDescription
anomaly_detectRun anomaly detection on a temporal stream
anomaly_subscribeSubscribe to real-time anomaly events
forecast_predictGenerate forecasts with confidence intervals
forecast_explainExplain forecast reasoning
pattern_detectDetect cycles and motifs in time series
pattern_summarizeSummarize detected patterns
temporal_enrichEnrich external data with temporal context
temporal_learnSubmit outcome data for model updates

Architecture at a Glance

ComponentRoleOTP Pattern
Anomaly EngineAdaptive SSM (Mamba-style) with selective state gatingGenServer per stream
Forecast EngineMulti-timescale CCO prediction, no forgettingGenServer + Nx
Pattern EngineSpectral + autocorrelation cycle detectionConcurrent tasks
Multi-Timescale StoreFast (ETS) → Medium → Slow → Glacial memoryTiered storage
Stream DashboardAnomaly timeline, forecast explorer, pattern viewerPhoenix LiveView

Multi-Timescale Memory

TierLatencyStorageContents
FastsecondsETSPer-tick state snapshots
MediumhoursSQLite/TimescaleDBAggregated state vectors
SlowdaysSQLite/TimescaleDBConsolidated summaries
GlacialmonthsSQLite/TimescaleDBLong-range seasonal models

Paired With

ProductRelationship
GeoFleeticSpatial intelligence — when + where = complete situational awareness
GraphonomousContinual learning substrate for temporal knowledge
DelegaticGovernance enforcement for temporal operations

Tech Stack

LayerTechnology
LanguageElixir 1.17+
FrameworkPhoenix 1.8+
DatabasePostgreSQL 16+ + TimescaleDB
Edge StorageSQLite
MLBumblebee / Nx (SSM, CCO models)
ReplicationEpoch-aware delta-CRDTs

Project Links

[&] Ampersand Box Design — ticktickclock.com

Open in the interactive atlas