Piramal Finance AI collections reach ₹754 crore monthly
Key highlights
- Piramal reported ₹754 crore of monthly hands-free collections in Q2 FY27, compared with ₹300 crore in Q2 FY26.
- Piramal generated ₹102 crore of cross-sell disbursals through conversational AI in Q2 FY27.
- Piramal’s fraud-decisioning AI gave 20,93,724 alerts in Q2 FY27, versus 10,85,760 a year earlier.
- Piramal said 74% of customer resolutions were serviced through automated WhatsApp, mobile-app and email channels in Q2 FY27.
powered by IRIS

In Q2 FY27, Piramal Finance reported ₹754 crore of monthly hands-free collections, up from ₹300 crore a year earlier, and ₹102 crore of quarterly cross-sell disbursals through conversational artificial intelligence (AI). The dashboard also recorded higher AI volumes in fraud checks, customer servicing, onboarding and internal workflows.
Know Piramal Finance inside out
Results, valuation, peers and who's buying, in one view.
How large were Piramal’s AI-linked collections and disbursals?
Piramal’s largest disclosed AI-linked monetary flow was ₹754 crore of monthly collections through hands-free collections in Q2 FY27, compared with ₹300 crore in Q2 FY26. Separately, Piramal recorded ₹102 crore of quarterly cross-sell disbursals through a conversational AI assistant in Q2 FY27; the prior-year dashboard did not disclose a comparative value for that metric.
The two measures cover different processes and time periods: hands-free collections are monthly collections, while conversational AI cross-sell is a quarterly disbursal figure. Piramal’s Q2 FY27 CX.ai spotlight also reported ₹172 crore of digital collections through Piramal Instant Assistant (PIA), covering overdue equated monthly instalment (EMI) and fee collections on WhatsApp and the Piramal App with real-time settlements.
Which underwriting and fraud AI processes expanded at Piramal?
Piramal increased AI activity in several underwriting and fraud-control workflows between Q2 FY26 and Q2 FY27. AI-assisted bank-statement processing rose to 1,94,486 statements from 1,82,551, while image intelligence processed 2,12,757 images for credit decisioning, compared with 59,362 a year earlier.
Document-fraud intelligence had the largest disclosed increase among these workflows, scanning 6,06,712 documents for tampering and fraud in Q2 FY27, up from 1,13,732 in Q2 FY26. Piramal’s fraud-decisioning AI engine gave alerts in 20,93,724 cases, compared with 10,85,760 cases a year earlier, although the presentation does not disclose prevented fraud amounts or resulting loan-performance outcomes.
Piramal also reported employee-facing use cases. Queries on ARYA, described as a work assistant, coach and friend, increased to 5,56,400 in Q2 FY27 from 3,85,000 in Q2 FY26, while AI-assisted contest-gamification messages rose to 2,11,600 from 1,68,000.
How did Piramal’s AI tools change customer servicing?
Piramal reported that 74% of customer resolutions were serviced through WhatsApp, mobile-app and email automation in Q2 FY27, up from 67% in Q2 FY26. The number of live voice and chat bots rose to 49 from 12, while bot-based onboarding accounted for 41% of total onboarding, compared with 11% a year earlier.
PIA recorded 6.3 lakh interactions in Q2 FY27, an increase of 54% year on year, according to the CX.ai spotlight. Piramal describes the self-service assistant as operating on WhatsApp and the Piramal App for customer queries, payments and new-loan offers, placing it in servicing, payment activity and customer acquisition.
Piramal said its voice-AI system handled about 5 lakh calls in Q2 FY27, double the previous quarter’s volume. The calls covered welcome calls, EMI reminders and promise-to-pay follow-ups, while the dashboard’s 49 live bots is the broader disclosed measure of voice and chat-bot availability.
Got a follow-up? Ask IRIS
What does Piramal’s CX.ai spotlight show about cross-sell?
Piramal said PIA generated about 29,000 cross-sell leads in Q2 FY27, up 26% quarter on quarter and representing 20% of total leads generated. PIA is Piramal’s in-house generative-AI conversational assistant, and management said it supports cross-sell from lead generation to disbursement.
The ₹102 crore of conversational-AI cross-sell disbursals was three times the prior quarter’s level, according to the Q2 FY27 spotlight. That result sits within a wider cross-sell franchise of 3.7 million customers, or 56.2% of Piramal’s 6.5 million total customer franchise, versus 2.9 million customers, or 55.5%, out of 5.2 million in Q2 FY26.
Cross-sell represented 30% of unsecured disbursements in both Q2 FY26 and Q2 FY27, while total unsecured disbursements increased to ₹6,339 crore from ₹3,776 crore. In total retail disbursements, the cross-sell share rose to 14% from 11%, as total retail disbursements increased to ₹14,186 crore from ₹10,619 crore.
What do Piramal’s AI metrics establish and not establish?
Piramal’s disclosures establish that AI tools were deployed across collections, cross-sell, fraud screening, document handling, servicing and onboarding in Q2 FY27. The reported measures include ₹754 crore of monthly hands-free collections, 20,93,724 fraud-alert cases, 74% automated-channel servicing and 41% bot-based onboarding.
The measures do not establish the profitability, credit-quality or loss-reduction effect of each tool. Piramal disclosed ₹102 crore of conversational-AI disbursals and 6,06,712 fraud-document scans, but did not provide AI-specific approval rates, delinquencies, fraud losses prevented, customer-retention outcomes or operating costs.
Conclusion
Piramal’s Q2 FY27 dashboard presents AI as an operating layer spanning collections, cross-sell, underwriting, fraud controls, customer resolutions and onboarding. The rise in monthly hands-free collections from ₹300 crore to ₹754 crore, together with ₹102 crore of conversational-AI cross-sell disbursals, coincided with higher processing volumes and broader automation coverage.
The unresolved issue is whether later dashboards will provide outcome measures beyond activity volumes and monetary flows. Piramal identified Q2 FY27 as its CX.ai spotlight, after Credit.ai in Q1 FY27, Collections.ai in Q4 FY26 and Operations.ai in Q3 FY26, but did not disclose an AI target, a scheduled next spotlight or AI-specific profitability and credit-performance results.


