Chat Assistant

Interact with AI for instant insights and assistance

Chat Assistant Overview

The Reclamia Chat Assistant lets you ask questions and get instant answers about your claims, operations, and procedures.

Accessing the Assistant

1. Click “Chat” in main navigation
2. Select agent (if multiple available)
3. Type your question
4. Get instant response

What You Can Ask

Claims Queries

Get specific claim info:

• “Show me claim #1234”
• “What claims are assigned to John?”
• “List all overdue claims”
• “Find quality defects from last week”

Analyze patterns:

• “Which category has most claims?”
• “Show me claims from the quality team”
• “What’s the average resolution time?”
• “How many claims in maintenance?”

SLA Queries

Check deadlines:

• “Which claims are at-risk?”
• “How many claims are overdue?”
• “When is claim #1234 due?”
• “What’s our SLA compliance?”

Workload status:

• “Is John overloaded?”
• “Which department is busiest?”
• “Show me team workload”
• “How many claims per person?”

Analytics Queries

Get insights:

• “What’s causing most quality issues?”
• “Are safety incidents increasing?”
• “Show me trends this month”
• “Which supplier has most issues?”

Performance review:

• “How’s the team doing?”
• “What’s our SLA compliance rate?”
• “Which department performs best?”
• “Show me weekly stats”

Procedure Queries

Get procedures:

• “How do I calibrate the press?”
• “What’s the conveyor maintenance schedule?”
• “Show me safety procedures for chemical spill”
• “What’s the customer complaint process?”

Training questions:

• “How do I troubleshoot hydraulic leaks?”
• “What’s the first aid procedure?”
• “Show me equipment startup checklist”
• “What’s our quality control standard?”

Asking Effective Questions

Be Specific

Good: “Show me high-priority quality claims from last 7 days” Poor: “Show me claims”

Provide Context

Good: “John has been handling conveyor issues. How many has he resolved?” Poor: “Tell me about John”

Use Natural Language

Good: “Which departments are struggling with SLA?” Poor: “dept sla ?”

Ask Follow-ups

Assistant: “Quality issues up 15% this month” You: “Why are they up? What categories increased?”

Understanding Responses

Citation Format

Responses include sources:

Q: "What's the conveyor maintenance schedule?"
A: "According to the Maintenance Schedule (last updated Jan 2025):
   - Oil changes: Monthly
   - Belt inspection: Quarterly
   - Complete rebuild: Every 2 years
   Source: maintenance_schedule.pdf"

Confidence Levels

Assistant indicates certainty:

• Certain: “Definitely” / “Per procedure” / “Historical data shows”
• Likely: “Probably” / “Appears to be” / “Trend suggests”
• Uncertain: “Can’t determine” / “Need more info” / “Not found in database”

Data Freshness

Assistant shows when data is current:

• “As of today:”
• “Latest data:”
• “Last 24 hours:”
• “Last updated: [date]“

Chat History

Viewing Conversation History

• Conversations auto-save
• Scroll up to see previous messages
• See all chat sessions in sidebar
• Search past conversations

Saving Important Conversations

• Star/favorite important exchanges
• Export conversation as PDF
• Share with team members
• Reference for documentation

Privacy

• Chat history only visible to you
• Can be seen by admins for compliance
• Retained per retention policy
• Can request deletion

Common Chat Scenarios

Scenario 1: Quick Daily Check

User: “What’s our status today?”

Assistant (provides):

• New claims count
• Claims resolved
• At-risk claims
• Overdue claims
• SLA compliance
• Top issues

Time saved: 10 minutes

Scenario 2: Troubleshooting

User: “Hydraulic pump making noise and losing pressure. What do I do?”

Assistant (provides):

• Step-by-step diagnosis
• Equipment manual reference
• Similar past cases
• Part specifications
• Estimated fix time

Time saved: 1 hour of trial-and-error

Scenario 3: Management Decision

User: “Should we hire more maintenance staff?”

Assistant (provides):

• Current workload metrics
• Maintenance claim trends
• Average resolution time
• Staffing efficiency ratio
• Historical hiring impact
• Recommendation

Time saved: Several hours of analysis

Scenario 4: Compliance Check

User: “Our safety compliance this quarter?”

Assistant (provides):

• Safety claims count
• Average resolution time
• SLA compliance
• Incident types breakdown
• Preventive actions taken
• Audit-ready report

Time saved: 30 minutes of manual compilation

Tips for Best Results

✅ Do:

• Ask clear, specific questions
• Provide context when needed
• Ask follow-up questions
• Review sources provided
• Save important answers
• Share helpful responses with team

❌ Don’t:

• Ask vague questions
• Expect perfect answers every time
• Skip reading sources
• Use for sensitive decisions without verification
• Rely solely on assistant for critical decisions
• Ignore “I’m not sure” responses

Limitations

Assistant cannot:

• Make binding decisions (you decide)
• Create new claims (use form)
• Delete data
• Modify procedures (admin only)
• Make purchases or commitments
• Replace human judgment

---

Next Steps

• Agent Configuration
• Agent Use Cases