Solace AI Event Connector
The Solace AI Event Connector is an important part of Solace Agent Mesh. It is a powerful tool designed to seamlessly integrate AI capabilities into your event-driven architecture. It enables you to create efficient pipelines that can process events from PubSub+ event brokers, using AI and other types of components, and then publish the results back to PubSub+ event brokers.
GitHub
For more information, see Solace AI Event Connector Github.
Key Features
- Event Processing Pipeline: Creates flows consisting of input components, processing components, and output components.
- AI Model Integration: Built on LangChain and LiteLLM supporting various AI models and providers.
- Extensible Architecture: Supports custom plugins and components for specialized processing needs.
- Resilient Processing: Ensures reliable event handling with acknowledgment-based processing.
- Scalable Design: Supports parallel processing and multiple component instances.
How It Works
Solace AI Event Connector operates through a series of interconnected components:
-
Flow Structure
- Each flow has one input component
- Multiple processing components
- One output component
- Components are connected via queues for buffering
-
Component Processing
- Each component runs in its own thread
- Processes one message at a time
- Supports input transforms and selections
- Can be scaled with multiple instances for parallel processing
- Event Management
- Tracks events throughout the processing pipeline
- Uses acknowledgment system for reliable delivery
- Ensures no event loss during processing
Configuration
The Solace AI Event Connector is configured through YAML files that allows you to:
- Define flow components and their sequence
- Set queue depths between components
- Configure the number of parallel instances
- Specify input/output parameters
- Define processing logic and transformations
- Passing the source path for the custom python components
For more information about the connector, see Solace AI Event Connector.