Analytics
Workspace dashboards, Auto-QA, AI insights, evaluations and simulations.
Filters:
Auto-refresh · 5m
Total conversations
24,847
12%vs previous 30d
AI resolution rate
67.2%
3.4ppfully self-served
Avg first response
1m 22s
18sacross channels
Avg full resolution
14m 47s
4mmedian 9m 12s
CSAT (post-survey)
4.42/5
0.08n=3,219 surveys
Predicted CSAT
4.38/5
0.06all conversations
Voice containment
71.3%
2.1ppno human handoff
Cost per resolution
₹37.20
4.1₹blended LLM + telephony
Conversation volume by channel
Stacked, last 30 days
Resolution rate
AI vs human, last 30 days
Sentiment trajectory
Daily, last 14 days
AI handoff reasons
Top 10 intents
Volume and resolution %
Card statement download
2841
92%
Late fee waiver
1903
64%
EMI conversion
1722
88%
Card block / hotlist
1498
96%
Cheque book request
1322
91%
Account balance
1180
99%
Dispute transaction
998
41%
KYC update
884
72%
Loan eligibility
712
78%
NEFT / IMPS issue
648
69%
Top KB articles cited
Citations & CSAT correlation
How to download card statement
1892
4.6
Late payment fee policy (Cards)
1311
4.1
EMI conversion eligibility
1187
4.5
Block / replace damaged card
982
4.7
KYC re-verification — DPDP
803
4.3
Dispute Visa / Mastercard txn
711
3.9
NEFT vs RTGS vs IMPS
612
4.4
Cheque bounce charges
488
3.8
Volume heatmap
Hour × day of week
0
3
6
9
12
15
18
21
Mon
Tue
Wed
Thu
Fri
Sat
Sun
Language distribution
Customer tier mix
SLA compliance
Within SLA window
94.2%
in SLA
Target: 90%
Cost breakdown
₹ this month
LLM₹142.8k
Voice STT/TTS₹88.2k
BSP (WhatsApp)₹41.6k
Telephony₹32.4k
Other₹14.9k
AI-generated insights
5 new
Resolution rate dipped 4.2% on Tuesday afternoon (2–6 PM IST)
Root cause: spike in card-related queries coinciding with delayed LLM provider P95 (2.4s → 4.1s). Cards Specialist v2 saw 38% of impact.
EMI calculator flow has highest CSAT — 4.8/5 across 240 uses
Promote in Front Desk routing? Currently only surfaced when intent is explicitly named.
Hindi conversations show 12% higher handoff rate than English
Mostly clarification loops in the 'Late fee waiver' procedure. Consider Hindi prompt rewrite.
Customer Aarav Joshi contacted 8× this week (typical 1–2)
Sentiment trending negative. Possible churn signal — consider retention offer.
73 conversations escalated due to tool failure on requestLateFeeWaiver
Backend returned 5xx in 11.2% of calls vs 0.4% baseline. Page integrations on-call.