The Call history tab in Kalimna is your command center for monitoring and analyzing every call your AI assistants handle. Whether you're running a clinic in Doha, a real estate office in Dubai, or a restaurant in Riyadh, having complete visibility into your call center performance is essential.
This comprehensive dashboard shows you total call volumes, success rates (objectives achieved), average call duration, per-call costs, and detailed information about each individual conversation—including recordings, transcripts, AI evaluations, and webhook data. For businesses in Qatar, UAE, Saudi Arabia, Kuwait, Bahrain, and Oman running Arabic and English AI phone agents, this transparency ensures you always know exactly how your AI call center is performing.
This guide will walk you through every aspect of the Call history tab: from understanding high-level metrics to diving deep into individual call recordings, transcripts, and evaluations. You'll learn how to monitor performance, identify issues, and continuously improve your AI agents based on real conversation data.
What You'll Learn:
- How to navigate to and understand the Call history tab
- What key metrics mean (total calls, objectives achieved, average duration)
- How to review the calls overview table
- How to filter and search through call history
- How to view individual call details and status
- How to listen to call recordings
- How to read and analyze call transcripts
- How to review AI evaluations and performance scores
- How to monitor costs and post-call webhooks
- Best practices for ongoing call monitoring
Watch the Call History Tutorial
Follow along with this step-by-step video to see the Call history tab in action:
💡 Tip: Watch the video while following the written steps below.
What the Call History Tab Shows You
The Call history tab is a comprehensive overview of all calls handled by your AI phone agents. It provides both high-level analytics at the top and detailed per-call information in the table below.
Key Metrics at the Top:
- Total Number of Calls: How many calls your AI assistants have handled in total or within a selected time period
- Percentage of Calls That Achieved the Objective: If you've set objectives for your assistants (e.g., "book an appointment," "qualify the lead"), this shows the success rate
- Average Call Duration: The mean length of all calls, helping you understand typical conversation times
These metrics give you an instant performance snapshot. If your objective achievement rate is low, you might need to refine prompts. If average duration is too high, you might need to streamline conversations.
🌍 Gulf Region Example:
A medical clinic in Kuwait City sees they handled 247 calls last week with an 87% objective achievement rate (appointments successfully booked) and 4.2-minute average duration. This tells them their Arabic/English AI receptionist is performing well, with most patients getting appointments booked efficiently.
Step 1 – Open the Call History Tab
Log into your Kalimna dashboard and navigate to Call history in the left-side navigation menu. Click it to open the call monitoring interface.
This tab displays calls from all your AI assistants by default, though you may be able to filter by specific assistants depending on the interface version.
Step 2 – Understand the Calls Overview Table
Below the high-level metrics, you'll see a table or list of recent calls. Each row represents one call and displays key information:
What's Shown for Each Call:
- Call Type Icon: Indicates whether this was a web call, phone call, or other type (often shown as an icon)
- Call Status: Whether the call completed successfully, failed, or had other outcomes
- Client Phone Number: The caller's number (if available)
- Duration: How long the call lasted (in minutes and seconds)
- Assistant Name: Which AI agent handled the call
- Campaign: If the call was part of an outbound campaign, this shows which one
- Timestamp: When the call took place (date and time)
- Carrier: The telephony carrier or provider used
This table lets you quickly scan recent activity and identify patterns, anomalies, or calls worth investigating further.
Step 3 – Filter and Search Calls
For businesses with high call volumes, filtering and search features help you find specific calls quickly.
Common Filters (if available):
- Date Range: View calls from today, this week, this month, or custom date ranges
- Assistant: Filter by specific AI agents if you have multiple
- Status: Show only successful calls, failed calls, or other statuses
- Duration: Find very short calls (often spam/bots) or unusually long calls
- Campaign: Filter by specific outbound campaigns
Use filters to quickly identify patterns:
- Very short calls (under 10 seconds) might be bots or hang-ups
- Failed calls might indicate technical issues
- Calls without objectives achieved might need prompt improvements
Step 4 – View Individual Call Details
Click on any call in the table to open the detailed call view. This is where you can access everything about that specific conversation.
Call Details Page Includes:
- Full call metadata: Date, time, duration, participants, status
- Recording: Audio file of the entire conversation (if recording was enabled)
- Transcript: Complete text transcript of what was said
- Evaluation: AI-generated quality assessment and objective achievement
- Cost breakdown: How much this call cost
- Variables extracted: Data collected during the call (if post-call actions were configured)
- Webhook status: Whether post-call webhooks fired successfully
Step 5 – Listen to Call Recordings
If you enabled call recording for your assistants, you can listen to the full conversation directly in the Call history interface.
Why Listen to Recordings:
- Quality assurance: Verify the AI sounds natural and professional
- Training: Identify where the AI could improve
- Troubleshooting: Understand why a call failed or didn't achieve its objective
- Compliance: Review calls for regulatory or legal requirements
- Customer disputes: Have proof of what was said if there's a disagreement
⚠️ Privacy Reminder:
Only enable call recording if it complies with local regulations in your region. Ensure you have proper consent disclosures and data protection policies in place. Review regulations in Qatar, UAE, Saudi Arabia, Kuwait, Bahrain, and Oman regarding call recording requirements.
Step 6 – Review Call Transcripts
Every call automatically generates a text transcript showing the complete conversation between the AI assistant and the caller. Transcripts are available even if you didn't enable recording.
What Transcripts Show:
- Full conversation flow: Exactly what both parties said
- Speaker labels: Clear indication of who spoke each line (AI vs. caller)
- Timestamps: When each part of the conversation happened
- Language detection: Whether the conversation was in Arabic, English, or mixed
Why Review Transcripts:
- Faster than listening: You can read a transcript much faster than listening to a recording
- Searchable: Find specific keywords or phrases across conversations
- Identify patterns: See common questions, objections, or confusion points
- Improve prompts: Notice where the AI gives unclear or incorrect responses
- Language quality check: For bilingual agents, verify Arabic and English responses are accurate
✨ Best Practice:
Review transcripts from your first 20-30 calls in detail, even if they seem successful. This helps you identify subtle issues early and refine your agent's prompts before scaling up. Pay special attention to how the AI handles Arabic/English code-switching for Gulf market customers.
Step 7 – Check Call Evaluations
For each call, Kalimna's AI generates an evaluation—an automated quality assessment of how well the conversation went.
What Evaluations Include:
- Objective Achievement: Did the call accomplish its goal? (e.g., appointment booked, lead qualified, question answered)
- Quality Score: Overall assessment of conversation quality
- Issues Identified: Potential problems like confusion, repetition, or incomplete information
- Summary: Brief recap of what happened in the call
How to Use Evaluations:
- Quick filtering: Identify low-scoring calls that need review
- Trend analysis: Track whether quality is improving or declining over time
- Training data: Use low-scoring calls to identify what to improve
- Celebrate wins: High-scoring calls show what's working well
Step 8 – Monitor Costs and Post-Call Webhooks
Per-Call Costs
Each call detail shows the exact cost for that conversation, calculated based on:
- Call duration (charged per minute)
- Features used (transcription, synthesis, LLM model, etc.)
- Any additional services (recordings, evaluations)
Monitoring costs helps you:
- Budget accurately for your AI call center
- Identify unusually expensive calls (might indicate inefficiencies)
- Calculate ROI compared to human receptionists
- Forecast monthly expenses based on call volume
Post-Call Webhooks Status
If you configured post-call actions and webhooks (to send data to your CRM, calendar, or other systems), the call details show:
- Webhook fired successfully: Data was sent and received by your endpoint
- Webhook failed: There was an issue (check your endpoint or contact support)
- Variables extracted: The actual data collected during the call (names, emails, appointment times, etc.)
This visibility ensures your integrations are working correctly and data is flowing to your business systems as expected.
Best Practices for Call Monitoring
1. Review Calls Regularly
Don't wait for problems to surface. Set aside time weekly to review:
- High-level metrics (are objectives being achieved?)
- A sample of recent calls (listen to 5-10 randomly)
- Any calls with low evaluation scores
- Failed or very short calls (under 30 seconds)
2. Track Trends Over Time
Monitor how metrics change:
- Is average duration increasing or decreasing?
- Is objective achievement improving with prompt refinements?
- Are costs per call staying consistent?
- Are there more calls at certain times of day?
3. Use Transcripts for Training
Transcripts are gold mines for improving your AI:
- Find common questions not in your Knowledge Base (add them)
- Identify where the AI gets confused (refine prompts)
- Notice successful conversation patterns (replicate them)
- Spot opportunities for new Mid Call Tools
4. Monitor Arabic and English Quality Separately
For bilingual agents serving Gulf markets:
- Review Arabic conversations to ensure dialect accuracy
- Check English conversations for clarity and professionalism
- Pay attention to code-switching (when customers mix languages)
- Verify both languages achieve similar objective rates
5. Share Insights with Your Team
Don't keep Call history data to yourself:
- Share successful call examples with sales/support teams
- Use transcripts to train human staff on common questions
- Report metrics to management to demonstrate ROI
- Identify product/service issues mentioned in calls
🌍 Gulf Business Insight:
A restaurant chain with locations in Dubai, Riyadh, and Kuwait City uses Call history to compare performance across cities. They notice Riyadh customers ask more questions in Arabic about halal certification, so they update their Saudi Arabia Knowledge Base with detailed halal information—improving objective achievement in that market.
What's Next?
You now know how to use the Call history tab to monitor, analyze, and improve your AI phone agents. You can review high-level metrics, dive into individual calls, listen to recordings, read transcripts, and verify that your integrations are working correctly.
Recommended Next Steps:
🛡️ Block Problem Callers
Use Call history to identify spam or bot callers, then add them to your Blacklist to prevent future waste.
View Guide →📚 Update Knowledge Base
Notice common questions in transcripts that your AI struggles with? Add that information to your Knowledge Base.
View Guide →🔧 Add Mid Call Tools
See opportunities in transcripts where real-time data lookups would help? Create Mid Call Tools to fetch that info.
View Guide →🤖 Refine Your Agent
Use insights from Call history to adjust your AI agent's settings, prompts, and behavior for better performance.
View Guide →Monitor Your AI Call Center with Confidence
Whether you run a clinic in Doha, a real estate office in Dubai, a restaurant in Riyadh, a hotel in Manama, or an e-commerce brand in Kuwait City or Muscat—Call history gives you full visibility into how your Arabic and English AI phone agents are performing, so you can continuously optimize and deliver better customer experiences.