Skip to main content

EventFlux Documentation

What is EventFlux?​

EventFlux is a pattern-first stream processing engine (CEP) built in Rust. It is designed for teams who need real-time computation without the complexity of a full streaming platform.

EventFlux enables you to compose filters, maps, joins, enrichment, windows, and temporal patterns, then execute them in milliseconds on-prem or in Kubernetes. With predictable performance, straightforward operations, and extensible connectors, EventFlux delivers the essentials of real-time computation—without platform lock-in or broker overhead.

The Name​

Event: We compute over events (not a broker/platform).

Flux: Continuous movement—data in motion, responsive by design.

Core Capabilities​

EventFlux provides essential stream processing capabilities:

  • Filters, maps, projections — Transform and filter event streams
  • Stateful and stateless operators — Build complex processing pipelines
  • Stream-to-table and stream-to-stream joins — Combine data from multiple sources
  • Enrichment via external stores and caches — Add context to events in real-time
  • Windows and watermarking — Tumbling, sliding, and session windows with late-data handling
  • CEP patterns — Sequence detection, absence patterns, correlation, and time-bounded triggers
  • Connectors — Extensible integration with common sources and sinks
  • Delivery guarantees — Exactly-once and at-least-once modes (where supported)
  • Observability — Metrics and tracing hooks for monitoring

Key Characteristics​

Pattern-First by Design​

Complex event patterns (sequences, absence, correlation) are built into the engine—not bolted on.

Rust-Fast, Predictable Latency​

Consistent performance with a small footprint, ideal for edge and on-prem deployments. No GC pauses or long warmups.

Essential, Not Excessive​

The stream processing you need—filters, joins, enrichment, windows—without platform bloat.

Operate Anywhere​

Clean self-hosted experience with Kubernetes-friendly deployment options.

Composable & Extensible​

Connectors for common sources and sinks. Pipelines are made of clear, reusable operators.

Who Uses EventFlux?​

EventFlux is designed for:

  • Platform and infrastructure teams that need real-time computing without running a full streaming platform
  • Fraud, risk, and security teams that require pattern detection and correlation
  • Observability and SRE teams correlating events across services and time windows
  • IoT and telemetry teams processing sensor data with enrichment at the edge

Technology​

EventFlux is built in Rust, providing:

  • Predictable latency with no garbage collection pauses
  • Memory safety with low overhead
  • Small binaries and fast startup times
  • Edge and on-prem friendly operations

Coming Soon​

EventFlux is currently under development. This documentation site will be populated with comprehensive guides, API references, and deployment instructions as the project progresses toward its 1.0 release.

Get Involved​

Visit eventflux.io for more information and updates.