Smart Suggestions

Smart Suggestions uses AI to recommend the best resolution for each claim, helping you make faster, more consistent decisions.

Types of Suggestions

Resolution Recommendations

Based on claim details, AI suggests:

• •Full Refund: Complete refund of purchase price
• •Partial Refund: Percentage refund for partial damage
• •Replacement: Send new product
• •Repair: Offer repair service
• •Store Credit: Issue credit for future purchase
• •Deny: Claim doesn't meet warranty criteria

Response Templates

AI drafts customer communications:

• •Approval messages
• •Information requests
• •Denial explanations
• •Follow-up messages

Assignment Suggestions

Recommends which team member should handle:

• •Based on expertise
• •Current workload
• •Claim complexity
• •Product category

How Suggestions Are Generated

The AI considers:

1. 1.Claim Value: Higher values may warrant different handling
2. 2.Product Category: Electronics vs clothing vs furniture
3. 3.Damage Type: Manufacturing defect vs user damage
4. 4.Customer History: Loyalty, previous claims, purchase value
5. 5.Similar Claims: How you resolved similar cases
6. 6.Policy Rules: Your configured warranty policies

Using Suggestions

Quick Actions

Click a suggestion to:

• •Apply the recommended resolution
• •Send the drafted response
• •Assign to suggested team member

Modifying Suggestions

You can:

• •Accept as-is
• •Edit before applying
• •Reject and choose different action
• •Provide feedback to improve AI

Confidence Indicators

Each suggestion shows:

• •Confidence percentage
• •Key factors considered
• •Similar past cases
• •Alternative options

Training the AI

Feedback Loop

Help the AI learn by:

• •Accepting accurate suggestions
• •Correcting inaccurate ones
• •Adding notes on why you chose differently

Building Your Model

Over time, the AI learns:

• •Your approval patterns
• •Your communication style
• •Your exception cases
• •Your team preferences

Best Practices

1. 1.Review before accepting: Always verify AI suggestions
2. 2.Be consistent: Similar claims should have similar outcomes
3. 3.Provide feedback: Help the AI improve
4. 4.Monitor quality: Check suggestion accuracy regularly
5. 5.Train your team: Ensure everyone uses suggestions the same way