990970: AI-Samples documentation

blazor/smart-ai-solutions/ai-samples/kanban/sentiment-analysis.md

@@ -228,9 +228,21 @@ await builder.Build().RunAsync();
 {% endhighlight %}
 {% endtabs %}
 
-## Razor Component
+## AI-powered Sentiment Analysis in Kanban
 
-This section demonstrates how to implement sentiment analysis in the Syncfusion Blazor Kanban component using AI. The AI Assistant evaluates the emotional tone of each task description and displays a corresponding emoji (😊, 😐, 😞) to help teams quickly assess the mood or urgency of tasks. This can be especially useful in agile workflows where emotional context can influence task priority and team communication.
+The AI-powered sentiment analysis feature in Syncfusion Blazor Kanban evaluates customer feedback for delivered tasks and displays emojis (😊, 😐, 😞) to indicate positive, neutral, or negative sentiment. This helps teams quickly assess satisfaction levels and prioritize follow-up actions, improving customer experience in service workflows.
+
+### UI Structure and User Interaction (`Home.razor`)
+
+The Razor page includes a Kanban board for managing pizza orders across columns like Menu, Order, Ready to Serve, and Delivered. A progress button triggers sentiment analysis, while custom card templates display task details such as title, description, and delivery date. After analysis, an emoji appears on delivered tasks to represent sentiment. A dialog template supports editing fields like category, size, and feedback, with conditional rendering based on task type.
+
+### AI Sentiment Analysis Logic (`Home.razor.cs`)
+
+The `GetScore` method serializes Kanban data to JSON and sends it to the AI service with a prompt requesting a `SentimentScore` (1–5) based on feedback. The AI response is cleaned and deserialized into the data model, updating each task’s score and assigning emojis: 😢 for 1–2, 😐 for 3, and 😀 for 4–5. State flags like `ShowScore` control rendering, while progress events (`Begin`, `End`) update button text during processing.
+
+### Data Binding and Error Handling
+
+The `SfKanban` binds to the `Pizza` list using `Category` as the `KeyField`, with dynamic columns from `columnData`. Card templates conditionally render sentiment emojis for delivered tasks. If AI analysis fails, the spinner state prevents premature interactions, and `StateHasChanged()` refreshes the UI after completion.
 
 (`Home.razor`)
 
@@ -520,10 +532,6 @@ If the AI service fails to return a valid response, the Kanban will display an e
 - **Network Issues**: Check connectivity to the AI service endpoint, especially for self-hosted Ollama instances.
 - **Large Prompts**: Processing large text inputs may cause timeouts. Consider reducing the prompt size or optimizing the request for efficiency.
 
-## Performance Considerations
-
-When handling large text content, ensure the Ollama server has sufficient resources (CPU/GPU) to process requests efficiently. For long-form content or batch operations, consider splitting the input into smaller segments to avoid performance bottlenecks. Test the application with your specific use case to determine optimal performance.
-
 ## Sample Code
 
 A complete working example is available in the [Syncfusion Blazor AI Samples GitHub repository](https://github.com/syncfusion/smart-ai-samples).

blazor/smart-ai-solutions/ai-samples/kanban/smart-task-suggestion.md

@@ -228,9 +228,27 @@ await builder.Build().RunAsync();
 {% endhighlight %}
 {% endtabs %}
 
-## Razor Component 
+## AI-powered Kanban Smart Card Generation
 
-This section demonstrates how to implement the Syncfusion Blazor Kanban component with AI-powered task generation. The AI Assistant analyzes the provided project description and automatically suggests relevant tasks, which are then displayed in the Kanban board.
+The AI-powered Kanban feature integrates Syncfusion Blazor Kanban with an AI service to automatically create structured task cards from user input. It combines a simple UI for project details with backend logic that generates tasks and updates the Kanban board dynamically.
+
+### UI Structure and User Interaction (`Home.razor`)
+
+The Razor page includes an input form for project details and a Kanban board for visualizing tasks. Users enter a project description and task count, then click Generate Tasks to start AI processing. A modal dialog ensures clean input for new projects, while conditional rendering switches between input and Kanban views. Cards display essential details like title, description, and story points with drag-and-drop support.
+
+### AI Task Generation Logic (`Home.razor.cs`)
+
+The `GenerateProjectTasks` method builds a prompt for the AI service, requesting tasks in strict JSON format with fields like Id, Title, Status, and Due Date. The response is cleaned and deserialized into `SmartSuggestionDataModel` objects, which are added to the Kanban data source. Progress button events provide feedback during generation, and default values like `Status="Open"` and `Color="#000000"` are applied to new tasks.
+
+### Data Binding and Kanban Configuration
+
+The `SfKanban` component binds to the `smartSuggestion` list, using `Status` as the `KeyField` to organize cards into columns (To Do, In Progress, Review, Done).  
+Card templates are customized via `KanbanCardSettings` to show essential task details with visual clarity.  
+A fallback `SfGrid` view allows tabular task management with inline editing, while a toggle button switches between Grid and Kanban layouts.
+
+### Error Handling and User Feedback
+
+If AI generation fails, a toast notification displays an error message without disrupting the workflow. Input validation ensures both project details and task count are provided before submission. State flags manage UI transitions and prevent issues during asynchronous operations.
 
 (`Home.razor`)
 
@@ -573,10 +591,6 @@ If the AI service fails to return a valid response, the Kanban will display an e
 - **Network Issues**: Check connectivity to the AI service endpoint, especially for self-hosted Ollama instances.
 - **Large Prompts**: Processing large text inputs may cause timeouts. Consider reducing the prompt size or optimizing the request for efficiency.
 
-## Performance Considerations
-
-When handling large text content, ensure the Ollama server has sufficient resources (CPU/GPU) to process requests efficiently. For long-form content or batch operations, consider splitting the input into smaller segments to avoid performance bottlenecks. Test the application with your specific use case to determine optimal performance.
-
 ## Sample Code
 
 A complete working example is available in the [Syncfusion Blazor AI Samples GitHub repository](https://github.com/syncfusion/smart-ai-samples).