Platform Architecture Overview

High-level map of Fluxion modules, runtimes, and extension points. Use this to select the right entry point and understand how pieces compose.

---

1. Layered architecture

Connect  →  Streaming Engine  →  Rule Engine
                  │                │
          └──────►│                │
                  ▼                │
              Fluxion Core  ◄──────┘
                    │
                  Enrich

Layer	Purpose
Fluxion Core	Deterministic aggregation engine (stages/operators), expression evaluator, metrics, registries.
Fluxion Connect	Source/sink connectors (Kafka, Event Hubs, MongoDB, custom SPI).
Fluxion Enrich	Operators that call external systems ($httpCall, $sqlQuery).
Streaming Engine	Always-on pipelines (fetch → transform → deliver → checkpoint).
Rule Engine	JSON/DSL-defined rule sets for decisioning and governance workflows.

---

2. Module checklist

Feature	Module	Reference
Aggregation runtime	fluxion-core	Core Overview
Connectors	fluxion-connect	Connect Overview
Enrichment operators	fluxion-enrich	Enrich Overview
Rule engine	fluxion-rules	Rule Engine Overview
Streaming engine	(contained in fluxion-core module today)	Streaming Overview

---

3. Selecting an entry point

Need	Start here
Run aggregation pipelines in-process	fluxion-core (see Usage Guide)
Ingest from Kafka/Event Hubs/Mongo	fluxion-connect + Streaming Engine
Enrich with HTTP/SQL lookups	fluxion-enrich operators
Declarative rule authoring/governance	fluxion-rules (Rule Engine)
Long-running stream processing	Streaming Engine (with Connect + Enrich)

---

4. Extending Fluxion

1. Add capabilities at the lowest layer.
2. New operator/stage → contribute to Fluxion Core.
3. New connector → implement SPI in Fluxion Connect.
4. New enrichment behaviour → extend Fluxion Enrich.
5. Expose through runtimes. Surface new operators/connectors to rule/streaming engines so users can adopt them immediately.
6. Document + test. Update docs/tests alongside the new contribution to keep assistants and developers aligned.

---

5. Integration patterns

Scenario	Modules involved	Notes
Request/response service with pipeline execution	Core	Cache parsed stages, use PipelineExecutor.
Manual approval workflow (Temporal)	Core + Rules + Workflow integration	See Temporal Bridge.
Real-time enrichment from Kafka to HTTP	Core + Connect + Enrich + Streaming	Use streaming pipeline with Kafka source and HTTP sink/operators.
Analytics batch job with JDBC source	Core + custom source	Extend streaming SPI or use executor in batch mode.

---

6. References & next steps

• Core Overview – runtime APIs and extension points.
• Connect Overview – built-in connectors and SPI.
• Enrich Overview – HTTP/SQL operators.
• Streaming Overview – pipeline lifecycle, error policies.
• Rule Engine Overview – declarative rule runtime.
• Usage Guide – end-to-end pipeline execution tutorial.
• Examples Gallery – ready-made pipelines to copy.

This overview keeps the module map clear; dive into the linked guides for deeper, LLM-ready documentation.