Moose Architecture

When you run moose-cli dev, Moose creates a complete development environment that combines your code with pre-configured infrastructure:

Moose organizes your data pipeline into five distinct infrastructure modules:

Module	Moose Code	Infrastructure	Default
Ingestion	Data Models	HTTP POST endpoints	Rust
Streaming	Data Models, Streaming Functions	Streaming topics & consumers	Redpanda
Database	Data Models, Blocks	OLAP tables & views	ClickHouse
Orchestration	Workflows	Task queue & workers	Temporal
Consumption	Analytics APIs	HTTP GET endpoints	Rust

Moose is designed to be highly customizable. You can control which modules are included in your deployment using moose.config.toml:

[features]
streaming_engine = true  # Include Redpanda for streaming
workflows = false        # Disable Temporal for orchestration

Warning:

By default, workflows are disabled. You must set workflows = true to enable the workflow engine.

Component	Purpose	Access
Web Server	Handles HTTP requests for data ingestion and APIs	localhost:4000
ClickHouse	Analytical database for data storage and querying	localhost:18123
Redpanda	Kafka-compatible message broker for streaming	localhost:19092
Temporal	Workflow orchestration engine	localhost:7233 (Server) / localhost:8080 (UI)
Redis	Caching and temporary storage	Internal only

All these components are automatically configured to work together, creating a complete development environment that mirrors a production deployment.