Page-level AI Prompts
Page-Level AI Prompts allow you to add intelligent, context-aware AI capabilities that span across an entire page rather than being limited to individual components or forms. This creates a more cohesive, intelligent user experience where AI can assist users throughout their interaction with your application.
What Are Page-Level AI Prompts?
Page-Level AI Prompts provide AI assistance at the page scope, allowing for:
- Page Assistants - AI helpers that understand the full context of the page
- Cross-Component Intelligence - AI that can reference data from multiple components
- Contextual Guidance - Smart help that adapts to what the user is doing
- Dashboard Insights - Auto-generated summaries pulling from multiple data sources
- Smart Navigation - AI-powered suggestions for where to go or what to do next
Key Difference: Unlike component-level AI that focuses on a single table or form, page-level AI has visibility across all elements on the page, enabling more comprehensive and contextual responses.
Types of Page-Level AI Features
1. Page AI Assistant
An always-available AI interface that users can interact with to get help, ask questions, or perform actions related to the current page.
Capabilities:
- Answer questions about data visible anywhere on the page
- Explain how to use features or complete tasks
- Provide recommendations based on page context
- Execute actions through natural language commands
- Guide users through multi-step processes
Display Options:
2. Dashboard Summaries
Automatically generated insights that analyze all data and components on a dashboard page.
Features:
- Auto-generate executive summaries of dashboard data
- Identify trends across multiple charts and tables
- Highlight anomalies or items requiring attention
- Provide recommendations based on visible metrics
- Update dynamically as filters or date ranges change
Example Use Cases:
- "This Month's Performance" - Summary of sales, revenue, and key metrics
- "Operations Overview" - Status across inventory, orders, and shipping
- "Team Dashboard" - Workload, performance, and priorities across team members
3. Smart Search
Natural language search that works across all data and content on the page.
Capabilities:
- Search across multiple tables and components simultaneously
- Understand intent behind queries ("show me urgent items")
- Filter and highlight results across the page
- Navigate to specific records based on natural language descriptions
4. Contextual Help System
AI-powered help that adapts to what the user is doing on the page.
Features:
- Detect what user is trying to accomplish
- Offer relevant tips and guidance
- Explain unfamiliar fields or options
- Provide step-by-step instructions for tasks
- Link to relevant documentation
5. Action Recommendations
AI suggests next best actions based on page content and user behavior.
Examples:
- "You have 3 overdue tasks - would you like to update them?"
- "New orders are waiting approval - review them now?"
- "Inventory is low on 5 items - generate purchase orders?"
- "This customer hasn't been contacted in 30 days - send follow-up?"
Setting Up Page-Level AI
Basic Configuration
Configuring the Page Assistant
- In AI Settings, select Configure Page Assistant
- Set the assistant's context and capabilities:
- Page Context - Describe what this page is for
- Available Actions - What can AI help users do?
- Data Sources - Which components AI can reference
- Tone - Formal, casual, technical, etc.
- Add custom prompts or instructions for the AI
- Configure the UI appearance and position
- Test with sample questions
Setting Up Auto-Generated Summaries
- Enable Auto-Generated Summary in page AI settings
- Define what should be summarized:
- All components on page
- Specific components only
- Key metrics from selected data sources
- Choose when summary is generated:
- On page load
- When filters change
- On user request (button/link)
- At scheduled intervals
- Select display location (top of page, dedicated section, modal)
- Customize summary format and style
Advanced Features
Multi-Component Context
Enable AI to understand relationships between different components on the page:
- Reference data from tables, charts, and forms together
- Provide answers that synthesize information across sources
- Identify correlations between different data sets
Example:
User asks: "Why are my sales down?"
AI analyzes: Sales table, inventory levels, customer feedback, and marketing campaigns
Response: "Sales are down 15% this month, correlating with lower inventory of your top 3 products and a 20% decrease in marketing spend."
User asks: "Why are my sales down?"
AI analyzes: Sales table, inventory levels, customer feedback, and marketing campaigns
Response: "Sales are down 15% this month, correlating with lower inventory of your top 3 products and a 20% decrease in marketing spend."
Personalized Experiences
AI adapts based on the logged-in user:
- Greet user by name
- Reference user's role and permissions
- Show relevant data and actions based on user context
- Remember previous interactions and preferences
Proactive Notifications
AI actively monitors page data and alerts users to important changes:
- New records that require attention
- Thresholds crossed (sales targets, inventory minimums)
- Anomalies or unusual patterns detected
- Opportunities identified in the data
Workflow Automation
Users can trigger complex workflows through natural language commands:
- "Send reminder emails to all overdue clients"
- "Create purchase orders for items below minimum stock"
- "Generate this week's performance report"
- "Schedule follow-ups with leads from last month"
Common Use Cases by Page Type
Dashboard Pages
AI Features:
- Auto-generated executive summary of all metrics
- Trend analysis across multiple charts
- Anomaly detection and alerts
- Drill-down recommendations
User Interactions:
- "What are my top priorities today?"
- "Explain the revenue drop in Q2"
- "Compare this month to last month"
- "What should I focus on this week?"
Detail Pages
AI Features:
- Smart suggestions based on record data
- Related record recommendations
- Historical context and trends for this record
- Next action guidance
User Interactions:
- "What's the history with this customer?"
- "What should I do next for this project?"
- "Find similar records"
- "Suggest tags for this item"
List/Table Pages
AI Features:
- Smart filtering and search
- Bulk action recommendations
- Data quality insights
- Pattern identification
User Interactions:
- "Show me items that need attention"
- "Which records are incomplete?"
- "Group these by priority"
- "What patterns do you see?"
Form Pages
AI Features:
- Smart field suggestions
- Data validation and cleanup
- Auto-complete based on context
- Duplicate detection
User Interactions:
- "Fill in standard values for new customer"
- "Check if this contact already exists"
- "Suggest a category for this item"
- "Validate this address"
Best Practices
Design & UX
- Make AI Discoverable - Clear button or interface to access AI assistant
- Provide Examples - Show sample questions users can ask
- Set Expectations - Explain what AI can and can't do
- Visual Feedback - Show loading states and processing indicators
- Accessibility - Ensure AI interface works with keyboard and screen readers
Performance
- Optimize Context - Only include necessary components in AI context
- Use Caching - Cache summaries that don't change frequently
- Lazy Loading - Load AI features only when needed
- Progressive Enhancement - Page works without AI if needed
Content & Prompts
- Clear Instructions - Provide AI with clear page context and purpose
- Consistent Tone - Maintain brand voice across all AI responses
- Helpful Errors - Guide users when AI doesn't understand
- Regular Updates - Refine prompts based on user feedback
Security & Permissions
- Data Access - AI respects all existing permission rules
- User Context - AI only sees data the current user can access
- Action Permissions - AI-suggested actions require appropriate permissions
- Audit Logging - All AI interactions are logged
- Rate Limiting - Prevent abuse with request throttling
Troubleshooting
AI Not Understanding Page Context
- Review and enhance page context description
- Ensure component IDs and names are descriptive
- Check that relevant components are included in AI scope
- Provide more detailed instructions in AI configuration
Inconsistent or Irrelevant Responses
- Narrow AI's scope to most relevant components
- Add more specific guidelines in system prompt
- Use pre-configured prompts for common questions
- Review logs to see what context AI is receiving
Performance Issues
- Reduce number of components in AI context
- Implement caching for auto-generated summaries
- Optimize data queries to load faster
- Consider lazy-loading AI features
High Token Usage
- Monitor which features consume most tokens
- Limit frequency of auto-generated summaries
- Use pre-configured prompts instead of free-form when possible
- Optimize system prompts to be more concise
Measuring Success
Track these metrics to understand how page-level AI is being used:
- Usage Frequency - How often is AI accessed per page view?
- Question Types - What are users most commonly asking?
- Response Quality - Are users getting helpful answers?
- Task Completion - Does AI help users complete tasks faster?
- Error Rate - How often does AI fail to understand or respond?
- User Satisfaction - Collect feedback on AI helpfulness
Future Enhancements
Capabilities coming soon to page-level AI:
- Voice interaction support
- Multi-page context (AI remembers across page navigation)
- Collaborative AI (multiple users interact with same AI session)
- Custom AI models trained on your specific app data
- Integration with external AI services and APIs
Examples
Executive Dashboard
Page: Company Dashboard
Components: Sales Chart, Revenue Table, Team Performance, Alerts
Page AI Configuration:
- Context: "Executive dashboard showing company-wide KPIs and metrics"
- Auto-Summary: Enabled (on page load)
- Page Assistant: Enabled (floating chat button)
- Smart Search: Enabled
Sample Auto-Summary:
"Today's Overview: Revenue is up 12% vs last month ($847K total). Sales team met 95% of targets.
3 critical alerts require attention. Top performer: Sarah Chen with $156K in closed deals."
Sample Assistant Interactions:
User: "What needs my attention?"
AI: "You have 3 critical alerts: 2 overdue customer renewals and 1 inventory shortage.
Additionally, the Q2 forecast is 8% below target."
User: "Show me Sarah's performance details"
AI: "Sarah Chen has closed 23 deals this month totaling $156,330. That's 145% of her target.
Her average deal size is $6,797, up from $5,200 last month. Would you like to see her
deal breakdown?"
Support Ticket Management
Page: Support Ticket Queue
Components: Active Tickets Table, Priority Chart, Team Workload, SLA Status
Page AI Configuration:
- Context: "Customer support ticket management and routing"
- Page Assistant: Enabled (side panel)
- Action Recommendations: Enabled
- Smart Search: Enabled
Sample Interactions:
User: "Summarize my queue"
AI: "You have 18 active tickets: 4 urgent, 9 high priority, 5 normal. Average response time
is 2.3 hours. 2 tickets are approaching SLA deadline. Most common issue: password resets (7)."
User: "Which tickets should I work on first?"
AI: "Priority recommendations:
1. Ticket #4892 - Critical billing issue, SLA expires in 45 minutes
2. Ticket #4877 - Enterprise client technical problem, 3 hours until SLA
3. Ticket #4901 - High priority feature request from strategic account
Would you like me to show these tickets?"
We'd love to hear your feedback.