Scaling insurance renewals with Asmita, a bilingual GenAI voicebot
Oriserve built and deployed Asmita to automate four-wheeler policy renewal outreach for Tata AIG General Insurance — reaching 8,662 customers through 250,000+ automated call attempts and renewing 59.78% of them.
About the client
Tata AIG General Insurance is a joint venture between the Tata Group and American International Group (AIG), offering motor, health, travel and other general insurance products across India. This engagement focused on one of its highest-volume, most time-sensitive workflows: renewal outreach for four-wheeler policies (Auto Secure Private Car Package) approaching expiry, across a T-45 to T-day window relative to policy expiry. Outreach runs fully automated and 24/7, opening in Hindi by default with a seamless switch to English.
The challenge
Renewal outreach at this scale is a race against the clock — and manual calling couldn't keep pace.
A hard, time-boxed window: every policy has a fixed T-45 to T-day renewal window — miss it and coverage lapses.
Manual calling doesn't scale — reaching every expiring policyholder meant a call-center workload that grew with policy volume.
Generic reminders don't handle price pushback, competitor quotes, agent preference or trust concerns — all common on renewal calls.
Non-standard requests like IDV changes or add-on edits traditionally required a human handoff, breaking the conversation.
Compliance and bilingual (Hindi/English) delivery had to be guaranteed on every call — hard at human-agent scale.
Without structured disposition tracking, there was limited visibility into why a call succeeded, stalled or needed a retry.
Beyond an off-the-shelf bot: the AI engineering we ran
This is not a scripted IVR or a generic LLM wrapper — it is a purpose-engineered conversation system built on live client data.
- Grounded, hallucination-controlled conversation. Every premium, NCB, IDV and coverage figure the bot speaks is pulled live from Tata AIG's policy data at call time — never invented.
- Deep rebuttal engineering. 25+ distinct objection paths — price, competitor quotes, trust, preferred agent, service complaints — each with a scripted empathetic recovery path back to the renewal goal.
- Real-time bilingual delivery. Conversation opens in Hindi and switches to English on request mid-call, with mandatory compliance disclosure embedded at the right point in every flow.
- Reliability and redial logic. Structured calling windows and up to 16 attempts per lead across the T-45 to T window, with disposition-based retry rules.
- Two-way platform integration. Bidirectional API exchange with Tata AIG's MoEngage platform pushes leads into the dialer and syncs call outcomes back in real time.
- Live resolution of non-standard requests. Function-calling into Tata AIG's IPDS APIs lets the bot resolve requests like IDV changes on the call itself, instead of routing to a human agent.
- Disposition and suppression engineering. ~28 structured dispositions align to Tata AIG's reporting taxonomy, and an hourly paid-customer feed keeps the calling list suppressed in near real time.
Results, measured end-to-end
| Funnel metric | Result | Measured against |
|---|---|---|
| Customers reached | 8,662 | November 2025 pilot window (T-45 to T) |
| Qualified leads generated | 27.47% | 2,343 qualified leads |
| Lead-to-sale conversion | 67% | 1,571 sales from 2,343 leads |
| Overall renewal rate | 59.78% | 5,178 renewals |
The solution we designed — Asmita, a single seamless conversation
End-to-end journey
1. Verify & greet — Right-party verification, then a personalized opening that names the policy, vehicle and exact expiry date. 2. Pitch & handle objections — Presents the NCB and discount-adjusted premium, then works through price, trust or competitor pushback in real time. 3. Resolve on the spot — For IDV, plan or add-on change requests, the bot calls Tata AIG's IPDS APIs live, recalculates the premium, and continues the same call. 4. Confirm & collect payment — Reads back policy and vehicle details, then sends a secure payment link by SMS, WhatsApp and email. 5. Log & sync — Every outcome is tagged to a structured disposition and pushed back to MoEngage for real-time suppression and reporting.
Bilingual experience
Conversation opens in Hindi by default and switches to English on request mid-call, with mandatory compliance disclosure embedded at the right point in every flow.
Integrated systems
Bidirectional API exchange with Tata AIG's MoEngage platform pushes leads into the dialer and syncs call outcomes back in real time. Function-calling into IPDS APIs resolves non-standard requests like IDV changes live on the call. An hourly paid-customer feed keeps the calling list suppressed in near real time.
Measurement built in
~28 structured dispositions align to Tata AIG's reporting taxonomy, and 9 non-STP scenarios are resolvable live via IPDS function-calling. Outcomes are measured end-to-end — customers reached, qualified leads, lead-to-sale conversion, and renewal rate, both overall and by sourcing channel.
Renewal outreach at our scale is one of the most time-sensitive parts of our business, and manual calling was never going to keep pace. Working with Oriserve, we've automated this process end-to-end and reached a renewal rate approaching 60% across more than 8,000 customers, fully autonomously.
TVishal Gupta · Tata AIG General Insurance
Why a generic solution would fail here
Price objections. A generic scripted IVR would have broken on the first price objection; deep rebuttal engineering with 25+ objection paths keeps the conversation moving toward renewal.
Non-standard requests. A generic bot would have stalled on any IDV, plan or add-on change request; live IPDS function-calling resolves these on the call itself.
Outcome visibility. A generic bot gives no visibility into why calls succeeded or failed; ~28 structured dispositions tied to Tata AIG's own reporting taxonomy provide it.
Compliance and language consistency. Guaranteeing a compliance disclosure and bilingual delivery on every call is hard at human-agent scale; both are embedded systematically into every flow here.
The takeaway
A generic scripted IVR could not have handled Tata AIG's renewal book — it would have broken on the first price objection, stalled on any IDV change request, and left the client with no visibility into why calls succeeded or failed. By grounding every conversation in live policy data, engineering deep rebuttal handling, and integrating directly with MoEngage and IPDS, this first proof-of-concept turned a manual, capacity-constrained process into a 24/7 automated engine — reaching 8,662 customers through 250,000+ attempts and renewing 59.78% of them.
Have a workload like this?
We'll pull the closest live production analog and walk you through how the numbers were reached.
Scaling insurance renewals with Asmita, a bilingual GenAI voicebot
Oriserve built and deployed Asmita to automate four-wheeler policy renewal outreach for Tata AIG General Insurance — reaching 8,662 customers through 250,000+ automated call attempts and renewing 59.78% of them.
About the client
Tata AIG General Insurance is a joint venture between the Tata Group and American International Group (AIG), offering motor, health, travel and other general insurance products across India. This engagement focused on one of its highest-volume, most time-sensitive workflows: renewal outreach for four-wheeler policies (Auto Secure Private Car Package) approaching expiry, across a T-45 to T-day window relative to policy expiry. Outreach runs fully automated and 24/7, opening in Hindi by default with a seamless switch to English.
The challenge
Renewal outreach at this scale is a race against the clock — and manual calling couldn't keep pace.
A hard, time-boxed window: every policy has a fixed T-45 to T-day renewal window — miss it and coverage lapses.
Manual calling doesn't scale — reaching every expiring policyholder meant a call-center workload that grew with policy volume.
Generic reminders don't handle price pushback, competitor quotes, agent preference or trust concerns — all common on renewal calls.
Non-standard requests like IDV changes or add-on edits traditionally required a human handoff, breaking the conversation.
Compliance and bilingual (Hindi/English) delivery had to be guaranteed on every call — hard at human-agent scale.
Without structured disposition tracking, there was limited visibility into why a call succeeded, stalled or needed a retry.
Beyond an off-the-shelf bot: the AI engineering we ran
This is not a scripted IVR or a generic LLM wrapper — it is a purpose-engineered conversation system built on live client data.
- Grounded, hallucination-controlled conversation. Every premium, NCB, IDV and coverage figure the bot speaks is pulled live from Tata AIG's policy data at call time — never invented.
- Deep rebuttal engineering. 25+ distinct objection paths — price, competitor quotes, trust, preferred agent, service complaints — each with a scripted empathetic recovery path back to the renewal goal.
- Real-time bilingual delivery. Conversation opens in Hindi and switches to English on request mid-call, with mandatory compliance disclosure embedded at the right point in every flow.
- Reliability and redial logic. Structured calling windows and up to 16 attempts per lead across the T-45 to T window, with disposition-based retry rules.
- Two-way platform integration. Bidirectional API exchange with Tata AIG's MoEngage platform pushes leads into the dialer and syncs call outcomes back in real time.
- Live resolution of non-standard requests. Function-calling into Tata AIG's IPDS APIs lets the bot resolve requests like IDV changes on the call itself, instead of routing to a human agent.
- Disposition and suppression engineering. ~28 structured dispositions align to Tata AIG's reporting taxonomy, and an hourly paid-customer feed keeps the calling list suppressed in near real time.
Results, measured end-to-end
| Funnel metric | Result | Measured against |
|---|---|---|
| Customers reached | 8,662 | November 2025 pilot window (T-45 to T) |
| Qualified leads generated | 27.47% | 2,343 qualified leads |
| Lead-to-sale conversion | 67% | 1,571 sales from 2,343 leads |
| Overall renewal rate | 59.78% | 5,178 renewals |
The solution we designed — Asmita, a single seamless conversation
End-to-end journey
1. Verify & greet — Right-party verification, then a personalized opening that names the policy, vehicle and exact expiry date. 2. Pitch & handle objections — Presents the NCB and discount-adjusted premium, then works through price, trust or competitor pushback in real time. 3. Resolve on the spot — For IDV, plan or add-on change requests, the bot calls Tata AIG's IPDS APIs live, recalculates the premium, and continues the same call. 4. Confirm & collect payment — Reads back policy and vehicle details, then sends a secure payment link by SMS, WhatsApp and email. 5. Log & sync — Every outcome is tagged to a structured disposition and pushed back to MoEngage for real-time suppression and reporting.
Bilingual experience
Conversation opens in Hindi by default and switches to English on request mid-call, with mandatory compliance disclosure embedded at the right point in every flow.
Integrated systems
Bidirectional API exchange with Tata AIG's MoEngage platform pushes leads into the dialer and syncs call outcomes back in real time. Function-calling into IPDS APIs resolves non-standard requests like IDV changes live on the call. An hourly paid-customer feed keeps the calling list suppressed in near real time.
Measurement built in
~28 structured dispositions align to Tata AIG's reporting taxonomy, and 9 non-STP scenarios are resolvable live via IPDS function-calling. Outcomes are measured end-to-end — customers reached, qualified leads, lead-to-sale conversion, and renewal rate, both overall and by sourcing channel.
Renewal outreach at our scale is one of the most time-sensitive parts of our business, and manual calling was never going to keep pace. Working with Oriserve, we've automated this process end-to-end and reached a renewal rate approaching 60% across more than 8,000 customers, fully autonomously.
TVishal Gupta · Tata AIG General Insurance
Why a generic solution would fail here
Price objections. A generic scripted IVR would have broken on the first price objection; deep rebuttal engineering with 25+ objection paths keeps the conversation moving toward renewal.
Non-standard requests. A generic bot would have stalled on any IDV, plan or add-on change request; live IPDS function-calling resolves these on the call itself.
Outcome visibility. A generic bot gives no visibility into why calls succeeded or failed; ~28 structured dispositions tied to Tata AIG's own reporting taxonomy provide it.
Compliance and language consistency. Guaranteeing a compliance disclosure and bilingual delivery on every call is hard at human-agent scale; both are embedded systematically into every flow here.
The takeaway
A generic scripted IVR could not have handled Tata AIG's renewal book — it would have broken on the first price objection, stalled on any IDV change request, and left the client with no visibility into why calls succeeded or failed. By grounding every conversation in live policy data, engineering deep rebuttal handling, and integrating directly with MoEngage and IPDS, this first proof-of-concept turned a manual, capacity-constrained process into a 24/7 automated engine — reaching 8,662 customers through 250,000+ attempts and renewing 59.78% of them.
Have a workload like this?
We'll pull the closest live production analog and walk you through how the numbers were reached.