sf-agent-framework

# Feedback Analyzer Utility - Agent Instructions ## Purpose This utility provides instructions for AI agents to generate comprehensive feedback analysis solutions for Salesforce implementations, enabling organizations to collect, analyze, and act on user feedback systematically. ## Agent Instructions ### When to Generate Feedback Analysis Generate feedback analysis components when: - User satisfaction needs measurement - Feature adoption requires feedback - Support tickets need sentiment analysis - Product improvements need user input - Training effectiveness needs assessment - Change management needs feedback loops - Customer experience needs monitoring ### Core Components to Generate #### 1. Feedback Collection Engine Generate components that: - Capture feedback from multiple channels - Standardize feedback formats - Categorize feedback types - Route feedback to appropriate teams - Track feedback status - Integrate with existing systems Key collection methods: - In-app feedback widgets - Email survey integration - Chatter post monitoring - Support case analysis - Meeting note extraction - Form submissions #### 2. Sentiment Analysis Engine Create sentiment analysis that: - Analyzes text for emotional tone - Scores feedback positivity/negativity - Identifies key themes and topics - Detects urgency indicators - Tracks sentiment trends - Generates sentiment reports #### 3. Feedback Processing Pipeline Implement processing that: - Validates feedback data - Enriches with metadata - Applies categorization rules - Triggers workflows - Generates notifications - Updates dashboards ### Configuration Requirements #### Custom Objects Create these objects: ```yaml Feedback__c: - Source__c (Picklist) - Type__c (Picklist) - Subject__c (Text) - Description__c (Long Text Area) - Sentiment_Score__c (Number) - Category__c (Multi-Picklist) - Priority__c (Picklist) - Status__c (Picklist) - Submitted_By__c (Lookup to User) - Submitted_Date__c (DateTime) - Resolution__c (Long Text Area) Feedback_Analysis__c: - Feedback__c (Master-Detail) - Analysis_Type__c (Picklist) - Score__c (Number) - Keywords__c (Long Text Area) - Themes__c (Multi-Picklist) - Recommendations__c (Long Text Area) - Analysis_Date__c (DateTime) Feedback_Action__c: - Feedback__c (Lookup) - Action_Type__c (Picklist) - Assigned_To__c (Lookup to User) - Due_Date__c (Date) - Status__c (Picklist) - Resolution_Notes__c (Text Area) ``` #### Configuration Metadata ```yaml Feedback_Config__mdt: - Channel__c (Text) - Active__c (Checkbox) - Processing_Rules__c (Long Text) - Routing_Rules__c (Long Text) - Notification_Settings__c (Long Text) - Sentiment_Threshold__c (Number) ``` ### Analysis Algorithms to Implement #### Sentiment Scoring ``` 1. Tokenize feedback text 2. Apply sentiment lexicon 3. Calculate base sentiment score 4. Apply modifiers (negation, intensity) 5. Normalize score (-1 to +1) 6. Categorize (Positive/Neutral/Negative) ``` #### Theme Extraction ``` 1. Extract key phrases 2. Apply topic modeling 3. Identify recurring themes 4. Calculate theme frequency 5. Rank by importance 6. Map to categories ``` #### Priority Calculation ``` Priority Score = (Sentiment Weight × 0.3) + (User Impact × 0.3) + (Frequency × 0.2) + (Business Impact × 0.2) ``` ### Implementation Patterns #### Real-time Processing Pattern 1. Capture feedback event 2. Queue for processing 3. Apply sentiment analysis 4. Categorize and route 5. Trigger notifications 6. Update dashboards #### Batch Analysis Pattern 1. Collect feedback batch 2. Pre-process and clean 3. Run analysis algorithms 4. Generate insights 5. Create summary reports 6. Schedule follow-ups #### Continuous Improvement Pattern 1. Track feedback outcomes 2. Measure resolution effectiveness 3. Identify improvement areas 4. Update processing rules 5. Refine algorithms 6. Report on trends ### Feedback Categories to Support #### Product Feedback - Feature requests - Bug reports - Performance issues - UI/UX suggestions - Integration needs - Enhancement ideas #### Process Feedback - Workflow efficiency - Approval bottlenecks - Communication gaps - Training needs - Documentation clarity - Support quality #### User Experience Feedback - Ease of use - Navigation issues - Performance perception - Visual design - Mobile experience - Accessibility ### Integration Requirements #### Salesforce Integration - Case creation for issues - Chatter feed monitoring - Email-to-feedback - Knowledge article updates - Idea management - Survey integration #### External Integration - Slack notifications - JIRA ticket creation - Teams alerts - Email digests - SMS for urgent items - Webhook support #### Analytics Integration - Einstein Analytics dashboards - Tableau visualizations - Power BI reports - Custom analytics APIs - Real-time monitoring - Predictive insights ### Dashboard Components to Generate #### Executive Dashboard Display: - Overall satisfaction score - Sentiment trend analysis - Top feedback categories - Resolution metrics - ROI of improvements - Comparative analysis #### Operational Dashboard Show: - Active feedback queue - Processing status - Assignment distribution - SLA compliance - Response times - Volume trends #### Team Dashboard Include: - Team-specific feedback - Action item tracking - Performance metrics - Improvement areas - Success stories - Training needs ### Best Practices to Implement 1. **Collection Best Practices** - Make feedback easy to submit - Provide multiple channels - Acknowledge receipt - Set clear expectations - Follow up on submissions 2. **Analysis Best Practices** - Use consistent scoring - Validate algorithms regularly - Consider context - Avoid bias - Maintain transparency 3. **Action Best Practices** - Respond promptly - Close the loop - Track outcomes - Share improvements - Celebrate successes 4. **Privacy and Security** - Anonymize when needed - Secure sensitive data - Respect preferences - Comply with regulations - Audit access ### Advanced Features to Consider 1. **AI-Enhanced Analysis** - Natural language processing - Emotion detection - Intent recognition - Predictive categorization - Auto-response suggestions 2. **Proactive Feedback** - Behavioral triggers - Smart timing - Contextual requests - Micro-surveys - A/B testing 3. **Feedback Gamification** - Reward participation - Recognition programs - Leaderboards - Achievement badges - Team competitions ### Error Handling Instructions Handle these scenarios: 1. Invalid feedback format 2. Processing failures 3. Integration timeouts 4. Duplicate submissions 5. Spam detection Recovery strategies: - Queue failed items - Manual review options - Retry mechanisms - Error notifications - Fallback processing ### Testing Requirements Generate test classes for: 1. Feedback collection 2. Sentiment analysis accuracy 3. Routing logic 4. Integration points 5. Dashboard calculations ### Reporting Capabilities Generate reports for: - Feedback volume trends - Sentiment analysis - Category distribution - Resolution rates - Response times - Impact assessment ### Success Metrics Track and measure: - Feedback submission rate - Response time - Resolution rate - Satisfaction improvement - Action completion - ROI of changes