Off
No Observability
Disable observability entirely for maximum throughput.
x
Streaming SQL with time, lineage, and observability built in.
ChronusQL processes event streams with SQL, windows, pattern recognition, realtime telemetry, and temporal ML.
Chronusphaera turns every query into a live execution view with metrics, traces, event lineage, windows, and model-training explainability.
Simple FilteringWindow and Event AugumentationMatch_Recognize for regular expression matchingMachine Learning Online ClassifierEvent LineageNode TopologyRealtime Telemetry
01
Write SQL for systems that evolve over time.
ChronusQL extends SQL with explicit event-time semantics, temporal windows, pattern recognition, and streaming machine learning. The language is designed for realtime event systems where time is a first-class concept rather than an afterthought.
Event-Time Processing
Use TIMESTAMP BY to define deterministic event time across engine, database, and UI.
Windowed Aggregation
Aggregate streams globally, by key, or inside deterministic event-time windows.
Pattern Recognition
Detect temporal sequences and relationships using MATCH_RECOGNIZE.
Streaming ML
Define features, delayed labels, and online classifiers directly in SQL.
02
See the query execute in realtime.
Chronusphaera is not a static dashboard. It continuously tails execution telemetry and updates timelines, graphs, inspectors, windows, and event relationships as new data arrives.
Every component shares interaction state, so selecting a node or event updates the entire workspace.
Event Flow over time
Execution Timeline
Pipeline Graph
Node Inspector
Event Inspector
Output Streams
03
SQL compiled into explainable execution topology.
ChronosQL transforms temporal SQL into a fully observable execution topology composed of operators such as windows, aggregates, filters, pattern recognizers, and ML stages.
Query clause mapping
WITH-scoped topology
Node relationships
Edge throughput ratios
Live node selection
04
Follow events through the query plan.
Chronusphaera aligns events directly to the logical query plan rather than displaying an unordered stream. You can visually follow each event as it moves across operators and windows.
window membership
operator transitions
pattern matches
parent/child relationships
event causality
temporal ML lifecycle
05
Understand what every node is doing.
Clicking a node reveals data flow, latency, counts, histograms, neighboring nodes, and SQL clause mapping without losing the connection back to the original query.
operator identity
input/output counts
latency percentiles
histogram view
SQL clause mapping
neighboring nodes
06
Drill into individual events and lineage.
event payload
UTC event timestamp
parent events
child events
operator path
window membership
Move from a point on the timeline into its payload, parents, children, window membership, and operator path. Trace mode adds sampled event-level visibility.
07
Streaming ML with visible training lifecycle.
ChronosQL can train online classifiers from streaming events while preserving event-time semantics and delayed-label causality.
online classifier
delayed labels
future-cause links
prediction confidence
training examples
active model window
08
Choose the observability level you need.
Metrics
Aggregated Observability
Capture node metrics, topology metadata, histograms, and percentiles for production monitoring.
Trace
Full Observability
Add sampled traces, lineage, windows, and event-level relationships for debugging and temporal analysis.
09
Built for understanding complex event systems.
Time Is Explicit
Event timestamps come from TIMESTAMP BY, preserving deterministic event-time behavior across runtime, storage, and UI.
Lineage Is Visible
Chronusphaera exposes parent/child event relationships across windows, operators, matches, and outputs.
Telemetry Is Adjustable
Move between throughput-oriented execution and deep debugging without changing the query itself.
10
Declarative stream processing. Inspectable execution.
ChronosQL makes temporal stream processing declarative. Chronusphaera makes it inspectable. Together they let you write, run, trace, and explain streaming systems from SQL down to individual events.