Workers Architecture
This section covers the foundational architecture of the worker system
Topics Covered:
- AI Workers: Conversational AI agents you can chat with
- AI Workflows: Visual automations built in Canvas
- How AI Workers and AI Workflows work together
- Lifecycle management (creation, deployment, versioning, sharing)
AI Workers
AI Workers are conversational AI agents designed to be flexible, modular, and reusable across various tasks. Users interact with AI Workers through a chat interface.
They are created using the AI Worker Builder, where Builders define:
- Profile: Name, avatar, tags, customizable welcome message
- Knowledge: Vector memory sources, context documents, semantic search
- Brain: LLM configuration, behavior prompts, role, tone, output format
- Skills: AI Workflows, API Providers (via Connectors), and MCP Tools
- Summary: Final review, visibility (public/private), and deployment
AI Workers support OAuth integrations, file uploads, and dynamic session contexts. They are ideal for users who need consistent, intelligent support across a range of workflows.
AI Workflows
AI Workflows are visual automations built for specific, structured tasks using the Canvas interface. Unlike AI Workers, users don't chat with AI Workflows — they execute them with defined inputs and receive structured outputs.
Key characteristics:
- Built visually: Drag-and-drop Canvas with node types (API calls, logic blocks, vector search, etc.)
- Composable: Can be embedded as "Skills" within AI Workers
- Modular: Each node can use its own connector and execute independently
- Flexible: Suitable for automation, backend processes, and domain-specific logic
How They Work Together
AI Workers and AI Workflows are complementary:
| AI Workers | AI Workflows | |
|---|---|---|
| Interface | Chat conversation | Structured input/output |
| Built with | AI Worker Builder | Canvas |
| Best for | Flexible, conversational tasks | Precise, repeatable automations |
Key relationship: AI Workflows can be added as Skills to AI Workers. This allows a conversational AI Worker to trigger precise automations when needed — combining the flexibility of chat with the reliability of structured workflows.
Lifecycle Management
-
Creation
- AI Workers: Built through a structured builder UI with 5 configuration tabs (Profile, Knowledge, Brain, Skills, Summary)
- AI Workflows: Built using Canvas, via manual configuration or AI-assisted Builder Chat
-
Deployment
- Workers are deployed into the Worker List (Launchpad) for Users to interact with
- Builders can set visibility (private/shared/public)
- AI Workflows can be embedded as Skills in other Workers
-
Versioning
- Workers are versioned implicitly through configuration history and deployment
- Future enhancements will include templating, rollback, and explicit version labels
-
Sharing
- AI Workers can be marked as public and shared across the organization
- AI Workflows are reusable as modules within other agents
- Templates can be saved and reused for consistent setups across teams
Summary
The EverWorker architecture is designed for modularity, scalability, and flexibility. AI Workers offer broad utility through conversational AI, while AI Workflows deliver precise automation through visual node-based design. Together, these components support a lifecycle that enables continuous creation, testing, reuse, and governance - all managed under a role-based platform that adapts to both technical and non-technical users.
Updated 16 days ago