Concepts
Core concepts and architecture behind VirtuousAI
Understanding VirtuousAI
Before diving into APIs and configurations, it helps to understand why VirtuousAI works the way it does. This section explains the core concepts that power the platform.
New to VirtuousAI? Start with Getting Started for a hands-on introduction, then return here to deepen your understanding.
What You'll Learn
| Concept | What It Covers |
|---|---|
| Chat Orchestration | How the AI assistant processes requests, uses tools, and maintains context |
| AI Agents | Composable AI workers with delegation, tool access, and specialization |
| Crystallization | Turning successful AI work into deterministic, repeatable automations |
| Data Pipeline | The bronze/silver/gold architecture for reliable data transformations |
| Data Modeling | Star schemas, fact/dimension tables, and why clean data matters |
Explore Concepts
Chat Orchestration
How AI conversations work: providers, context, and tool execution
AI Agents
Composable AI workers with delegation and specialized tool access
Crystallization
Convert successful AI operations into repeatable automations
Data Pipeline
Bronze, silver, and gold layers for progressive data refinement
Data Modeling
Star schemas, OLAP patterns, and dimensional modeling basics
How Concepts Connect
These concepts work together:
- Chat lets you ask questions about your data naturally
- Agents give AI specialized knowledge and composable delegation
- Crystallization captures successful AI workflows as repeatable automations
- Data Pipeline ensures your data is clean and queryable
- Data Modeling structures data for fast, meaningful analytics
When you ask the assistant "What were my top products last month?", it queries gold-layer tables built from cleaned silver data extracted from your raw bronze sources. When an agent successfully builds a report, you can crystallize that workflow into an automation that runs on a schedule.