{"id":72724,"date":"2026-09-08T16:14:56","date_gmt":"2026-09-08T10:44:56","guid":{"rendered":"https:\/\/newswiredelhi.com\/index.php\/2026\/09\/08\/arrowhead-ai-voice-agents-now-handle-40-of-inbound-customer-support-calls-on-kisshts-ring-app\/"},"modified":"2026-09-30T02:01:13","modified_gmt":"2026-09-29T20:31:13","slug":"arrowhead-ai-voice-agents-now-handle-40-of-inbound-customer-support-calls-on-kisshts-ring-app","status":"publish","type":"post","link":"https:\/\/newswiredelhi.com\/index.php\/2026\/09\/08\/arrowhead-ai-voice-agents-now-handle-40-of-inbound-customer-support-calls-on-kisshts-ring-app\/","title":{"rendered":"Arrowhead AI Voice Agents Now Handle 40% of Inbound Customer Support Calls on Kissht\u2019s Ring App"},"content":{"rendered":"

Within two months of deployment, the AI agents achieved an 88% automated resolution rate across selected support queries while reducing average handle time by approximately 30%<\/strong><\/em><\/p>\n

Bengaluru (Karnataka) [India], September 8:<\/strong> Arrowhead,<\/a><\/strong> a Bengaluru-based enterprise voice AI company, has announced results from its deployment with digital lending platform Kissht. Within two months of launch, Arrowhead\u2019s AI voice agents began handling 40% of all inbound customer support calls received through Kissht\u2019s Ring app.<\/p>\n

The deployment currently covers loan application rejection queries and requests related to mandate or No Objection Certificate cancellations. Approximately 88% of these calls are resolved without human intervention.<\/p>\n

\u201cWithin the first two months, the bot was handling 40% of our entire inbound volume,\u201d said Suraj Shetty, Head of Customer Experience, Learning and Development at Kissht (Ring). \u201cOf those calls, around 88% are getting handled without needing a human at all.\u201d<\/em><\/p>\n

Unlike many voice AI deployments that begin with outbound sales or engagement use cases, Arrowhead and Kissht introduced the technology directly into inbound customer support. The initial deployment focused on queries that can involve customer frustration, sensitive financial information and regulatory considerations.<\/p>\n

A Focused Deployment With Defined Guardrails<\/h3>\n

The AI agents were initially trained to handle two specific categories of calls: queries concerning rejected loan applications and requests involving mandate or NOC cancellations.<\/p>\n

The scope of the deployment was kept deliberately focused. The AI agent operates only within these predefined call categories. If a customer refers to the Reserve Bank of India during a conversation, the call is immediately transferred to a human specialist without the AI attempting to respond further.<\/p>\n

Arrowhead\u2019s<\/a> <\/strong>technology has also been integrated with Kissht\u2019s dialler and customer relationship management system. This allows relevant customer information and application status to be retrieved before the call is connected. As a result, customers are not required to repeatedly verify basic information or explain the same issue after a transfer.<\/p>\n

One of the principal technical challenges during development was reducing the delay between conversational turns.<\/p>\n

\u201cWhen we started, latency was around one second. We built our own small language model and brought that down to around 500 milliseconds, which is among the lowest in the industry,\u201d<\/em> said Devyani Gupta, Founder of Arrowhead.<\/p>\n

Reducing this delay was particularly important for creating a more natural interaction during sensitive customer conversations. Faster response times help reduce interruptions, awkward pauses and the impression that the customer is interacting with a conventional automated system.<\/p>\n

Approximately 30% Lower Average Handle Time<\/h3>\n

According to operational data from Kissht, the AI voice agents are completing calls approximately 30% faster than human specialists handling comparable queries.<\/p>\n

\u201cThe bot\u2019s average handle time is running about 30% lower than our human agents,\u201d said Shadab, Product and Build Owner at Kissht.<\/p>\n

\u201cIf the same cost per minute is assumed, that translates into an approximately 30% saving on every call resolved entirely by the AI agent,\u201d <\/em>Gupta added.<\/p>\n

The reduction in average handle time is in addition to the operational benefits created by resolving approximately 88% of calls within the selected categories without human involvement.<\/p>\n

Kissht has stated that the efficiency improvement has not resulted in a decline in the customer experience measured during the deployment.<\/p>\n

\u201cThat is what I would want another CX head to understand,\u201d Shetty said. \u201cWe reduced handle time, but the customer experience held steady.\u201d<\/em><\/p>\n

Extending Support Beyond Regular Working Hours<\/h2>\n

The deployment has also allowed Kissht to address customer calls received outside the working hours of its support specialists.<\/p>\n

Approximately 20% of the company\u2019s inbound support volume is received between 8 PM and 8 AM. Before the introduction of the AI voice agents, many of these calls remained in the queue until the following working day.<\/p>\n

\u201cAround 20% of our inbound volume comes between 8 PM and 8 AM. Earlier, that became a queue waiting for the next day,\u201d<\/em> Shetty said.<\/p>\n

Arrowhead\u2019s AI agents now provide round-the-clock coverage for the supported query categories. The same operational policies, escalation rules and guardrails apply to calls received during and outside regular support hours.<\/p>\n

A Collaborative Implementation Approach<\/h3>\n

Arrowhead and Kissht attributed the speed of the rollout to close collaboration between their respective product, customer experience, and engineering teams.<\/p>\n

The implementation process included daily stand-up meetings from the beginning of the engagement. Arrowhead\u2019s team also reviewed live customer conversations to understand how callers described their concerns, where conversations became difficult, and when human intervention was necessary.<\/p>\n

Kissht\u2019s implementation team worked directly with Arrowhead\u2019s founders and engineers throughout the deployment. This allowed feedback from customer calls to be incorporated into the system without relying solely on periodic reviews or conventional support-ticket processes.<\/p>\n

The two companies used this feedback to refine conversational flows, improve response speed, and strengthen the rules governing transfers to human specialists.<\/p>\n

EMI and Repayment Support Agent Under Development<\/h3>\n

Following the initial deployment, Arrowhead and Kissht are developing an AI agent for EMI and repayment-related queries.<\/p>\n

The next phase is expected to cover payment confirmations, auto-debit outcomes, duplicate deductions, late-fee status, foreclosure requests, NOC timelines, and CIBIL-related timelines.<\/p>\n

Because these conversations involve transaction-specific information, the new agent is being developed with stricter controls. One of its core instructions is described as \u201cmoney facts, absolute,\u201d meaning the system must not independently calculate, estimate or round financial information.<\/p>\n

The expansion represents a significant progression from handling defined service queries to managing conversations in which an inaccurate response could directly affect a customer\u2019s understanding of a payment or financial obligation.<\/p>\n

Kissht said the decision to proceed with the repayment use case followed the operational performance of the first deployment and the ability of the system to maintain customer experience while reducing handling time.<\/p>\n

Deployment Highlights<\/h3>\n
\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n
Metric<\/strong><\/td>\nResult<\/strong><\/td>\n<\/tr>\n
Share of total inbound volume handled by month two<\/td>\n40%<\/td>\n<\/tr>\n
Calls resolved without human involvement<\/td>\nApproximately 88%<\/td>\n<\/tr>\n
Average handle time compared with human specialists<\/td>\nApproximately 30% lower<\/td>\n<\/tr>\n
Estimated savings on fully automated calls, assuming the same cost per minute<\/td>\nApproximately 30%<\/td>\n<\/tr>\n
Customer experience<\/td>\nHeld steady<\/td>\n<\/tr>\n
Conversational turn latency<\/td>\nApproximately 500 milliseconds, reduced from around one second<\/td>\n<\/tr>\n
Inbound volume received between 8 PM and 8 AM<\/td>\nApproximately 20%<\/td>\n<\/tr>\n
Support availability<\/td>\n24\/7<\/td>\n<\/tr>\n
Call direction<\/td>\n100% inbound<\/td>\n<\/tr>\n
Current use cases<\/td>\nLoan rejection queries and mandate or NOC cancellation requests<\/td>\n<\/tr>\n
Use case under development<\/td>\nEMI and repayment support<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/figure>\n

About Kissht<\/h3>\n

Kissht is an Indian digital lending platform that provides consumer credit through its Ring app. Its offerings include personal loans, business loans and loans against property, with a focus on serving India\u2019s underserved and emerging middle-class consumers.<\/p>\n

About Arrowhead<\/h3>\n

Arrowhead develops enterprise-grade conversational AI voice agents for organisations managing customer interactions at scale. Headquartered in Bengaluru and backed by Stellaris Venture Partners, the company primarily works with banking, financial services and insurance organisations across inbound customer support and outbound engagement. Arrowhead develops its own low-latency speech and language technology for multilingual conversations in India.<\/p>\n

Media Contact<\/strong><\/p>\n

Garv Jain
<\/strong>Founder\u2019s Office, Arrowhead
Email: garv@arrowhead.team
Phone: +91 96095 21113
Website: arrowhead.ai<\/a><\/strong><\/p>\n

If you object to the content of this press release, please notify us at pr.error.rectification@gmail.com. We will respond and rectify the situation within 24 hours.<\/em><\/p>\n