AI Features — Walkthrough Guide
Overview
Vantage has AI deeply integrated throughout the platform. From the global AI assistant to per-tile summaries, query generation, and workflow AI nodes, AI helps you understand your data faster and make better decisions.
Global AI Assistant
The Global AI Assistant is accessible from anywhere in the platform via the AI icon in the sidebar or the keyboard shortcut.
What It Can Do
- Answer questions about your data: "What were our top 5 customers last quarter?"
- Generate queries from natural language: "Show me revenue by region for 2024"
- Explain data currently visible on your dashboard
- Suggest actions based on trends and anomalies
- Help with configuration: "How do I set up a Slack integration?"
How to Use It
- Click the AI icon in the sidebar or press the keyboard shortcut.
- The AI panel slides open on the right.
- Type your question or request in natural language.
- The AI responds using your data context, AI context snippets, and the currently visible dashboard.
Context Awareness
The AI assistant is context-aware — it knows:
- Which page you're on (dashboard, settings, workflows, etc.)
- Which tiles are visible and what data they contain
- Your AI context snippets (company overview, industry, custom instructions)
- Your preferred AI provider and model
Tile AI Features
Each dashboard tile has built-in AI capabilities:
Tile Summaries
AI-generated plain-language explanations of what the data in a tile shows.
- Automatically generated when a tile loads.
- Updated when data changes.
- Can be refreshed manually.
Popup AI Chat
Click into a tile to open a focused AI chat about that specific data:
- Click a tile to expand it.
- Click the AI chat icon in the popup editor.
- Ask questions about this specific tile's data.
- The AI responds with insights specific to this dataset.
AI Providers
Vantage supports multiple AI providers. Configure your preference in Settings → AI Features → Intuidy AI:
| Provider | Models |
|---|---|
| Intuidy AI | Default AI (no setup required) |
| OpenAI | GPT-4o, GPT-4o-mini, etc. |
| Claude (Anthropic) | Claude 3.5 Sonnet, Claude 3 Opus, etc. |
| Gemini (Google) | Gemini Pro, Gemini Ultra, etc. |
| DeepSeek | DeepSeek Chat, DeepSeek Coder |
| Grok (xAI) | Grok-1, Grok-2 |
| Mistral | Mistral Large, Mistral Medium, etc. |
To switch providers:
- Go to Settings → AI Features → Intuidy AI.
- Select your preferred provider.
- Select the specific model.
- Click "Save AI Settings".
AI Context
AI context defines the background information that the AI uses when generating responses.
Organization-Level Context (Snippets)
Set during onboarding or in Settings → AI Features → Context:
- Company Overview — What your company does.
- Industry & Market — Your industry for relevant benchmarks.
- Company Website — Your website for reference.
Custom Instructions
Set in Settings → AI Features → Context:
- Free-text instructions appended to every AI request.
- Examples: "Always format numbers with commas", "Focus on quarter-over-quarter trends", "Respond concisely".
AI Query Settings
Set in Settings → AI Features → Query Settings:
- Process Large Datasets — When enabled, AI processes all rows (slower but more accurate). When disabled, AI samples data (faster but may miss edge cases).
AI in Workflows
Workflows have dedicated AI nodes for automated data processing:
| Node | Purpose |
|---|---|
| AI Enrichment | Add AI-generated columns to your data (sentiment, category, summary, etc.) |
| AI Formatter | Reformat data using AI (clean text, standardize formats) |
| AI Summary | Generate summaries of data batches |
| AI Compliance Check | Evaluate rows against compliance policies (PASS/FAIL) |
| AI Conditional | Route data based on AI evaluation of conditions |
| AI Transcriber | Transcribe audio/video content |
| Image Analysis | Extract information from images |
| PDF Extract | Extract text and data from PDF files |
| Text Extraction | Extract structured data from unstructured text |
| AI Trigger | Start workflows based on AI-detected events |
Token Usage
All AI operations consume tokens. Monitor your usage in Settings → Account → Usage & Tokens.
Token costs vary by:
- Provider and model — More powerful models cost more tokens per request.
- Data volume — Processing more rows costs more tokens.
- Feature type — Workflow AI nodes, tile summaries, and assistant queries each have different costs.