Features Guide
Complete guide to orckAI's AI orchestration capabilities
Understand each feature, when to use it, and how they work together to create powerful AI-driven solutions. From simple chatbots to complex multi-step workflows.
Core Features
AI Agents
What It Does
AI Agents are intelligent assistants that can reason, use tools, and accomplish complex tasks autonomously. They combine language models with external capabilities to perform sophisticated operations.
Key Features
- Multiple LLM support (Claude, OpenAI, local models)
- Tool integration (web search, file operations, custom MCP tools)
- Extended thinking mode for complex reasoning
- Real-time streaming responses
- Knowledge base integration with citations
- Conversation memory and context retention
When to Use
- Interactive customer support requiring reasoning
- Complex data analysis and research tasks
- Document processing with multiple steps
- Technical troubleshooting and diagnostics
- Content creation with fact-checking
Workflows
What It Does
Workflows automate multi-step processes with triggers, conditions, and agent execution. Perfect for background processing and automated business operations.
Key Features
- Multiple trigger types (manual, schedule, file drop)
- Conditional logic and branching
- Variable interpolation between steps
- Agent step execution with context passing
- File processing and transformation
- Queue-based reliable execution
When to Use
- Scheduled report generation
- Document processing pipelines
- Data synchronization between systems
- Automated content moderation
- File-triggered processing workflows
Knowledge Bases
What It Does
Knowledge Bases provide AI agents with access to your documents and data, enabling accurate responses with proper citations and source attribution.
Key Features
- Multiple file format support (PDF, Word, Excel, text)
- Vector embeddings for semantic search
- Automatic citation generation
- Real-time document indexing
- Organization-scoped access control
- Chunk-based retrieval for context
When to Use
- Company policy and procedure queries
- Technical documentation assistance
- Legal document research
- Product information lookup
- Training material reference
MCP Servers
What It Does
MCP (Model Context Protocol) Servers connect AI agents to external systems, databases, and APIs, enabling real-time data access and system integration.
Key Features
- Docker-based server deployment
- Custom tool generation from specs
- Database connection management
- API integration capabilities
- Secure credential handling
- Tool discovery and registration
When to Use
- CRM system integration
- Database query automation
- Third-party API connections
- Legacy system modernization
- Real-time data retrieval
Prompt Templates
What It Does
Prompt Templates define reusable system prompts for consistent AI behavior across agents and widgets, ensuring brand voice and quality standards.
Key Features
- Reusable prompt definitions
- Citation format specification
- Knowledge base association
- Token limit configuration
- Brand voice consistency
- Multi-language support
When to Use
- Standardizing AI responses across channels
- Maintaining brand voice consistency
- Creating specialized expert personas
- Ensuring compliance requirements
- Template-based widget deployment
Widgets
What It Does
Widgets are embeddable AI chat interfaces that bring AI capabilities directly to your website or application with minimal integration effort.
Key Features
- Single-line JavaScript integration
- Customizable appearance and branding
- Prompt template integration
- Knowledge base access
- API key authentication
- Document download capabilities
When to Use
- Website customer support chat
- Product documentation assistance
- Lead qualification systems
- Internal help desk portals
- E-learning support interfaces
Advanced Capabilities
Multi-LLM Support
- Claude (Anthropic) integration
- OpenAI GPT models
- Local model deployment
- Model switching per agent
- Cost optimization strategies
Multi-Tenant Architecture
- Organization-scoped data isolation
- Row-level security
- Separate storage paths
- User access management
- Billing and usage tracking
Streaming Support
- Real-time response streaming
- WebSocket connections
- Progressive response display
- Interrupt and resume capability
- Low-latency interactions
Extended Thinking (Super Mode)
- Enhanced reasoning capabilities
- Complex problem decomposition
- Multi-step analysis
- Longer context processing
- Advanced tool orchestration
Feature Comparison Matrix
Compare the core features to choose the right approach for your use case:
| Capability | AI Agents | Workflows | Widgets |
|---|---|---|---|
| AI Reasoning | Full conversational AI | Agent step execution | Template-based responses |
| Tool Access | All tools available | Via agent steps | Limited to KB/search |
| Multi-step Logic | Autonomous reasoning | Explicit workflow steps | Single interaction |
| Scheduling | Manual execution | Cron scheduling | User-triggered only |
| User Facing | Interactive chat | Background processing | Embedded chat |
| KB Integration | Full access with citations | Via agent steps | Template-configured |
Choosing the Right Feature
Use AI Agents when: You need interactive, conversational AI with reasoning capabilities and tool access.
Use Workflows when: You need automated, scheduled, or triggered multi-step processes.
Use Widgets when: You need to embed AI chat on websites with specific branding and prompts.
Start Simple
Begin with Widgets for basic chat, then evolve to Agents for complex interactions, and Workflows for automation.
Combine Features
Use Knowledge Bases across all features, Prompt Templates for consistency, and MCP Servers for system integration.
Scale Gradually
Add features incrementally as your use cases evolve from simple Q&A to complex AI-driven automation.
Ready to Explore Features?
Start with a simple Widget and explore how features work together.