The Challenge
Ring lets a borrower compare loan offers from multiple NBFC lenders, each with its own amount, EMI, and processing fee, and then hit Proceed to move into the application. It is the highest-intent moment in the whole journey, and also the easiest place to lose someone.
Borrowers landed on the Offers page and saw EMIs with no explanation. Questions about lenders, fees, and approval went unanswered. Many were weighing three or more offers at once with no simple way to compare them side by side. Confused borrowers simply dropped off, and these were warm, ready-to-borrow leads walking away at the final step.
Ring needed a way to answer these questions instantly, in plain language, right on the page, without throwing more support agents at a problem that was happening at scale.
The Solution
RevRag AI deployed its in-app agent, Riya, directly on Ring's Available Loan Offers page. Borrowers could ask her anything about their offer and get a clear answer in real time. No callbacks, no queues, no leaving the page.
Riya was trained on Ring's offer logic, lender details, EMI and fee math, and the questions borrowers actually ask. She explains the numbers, helps borrowers compare lenders, nudges them toward the offers that fit, and walks them to the Proceed button with confidence. She also captures intent signals from every conversation, so Ring's product and CX teams can see exactly where borrowers hesitate. Throughout the pilot, response latency stayed comfortable and never got in the way.
Real-Time Query Resolution
Borrowers asked about EMIs, fees, lenders, and approval and got accurate answers on the spot. No drop-off while waiting for a callback.
Plain-Language Money Talk
Riya explains EMIs, lender differences, and processing fees in simple terms, the math borrowers were quietly doing in their heads before committing.
Multi-Lender Guidance
With several offers on screen, Riya guides borrowers to compare options and select multiple lenders before they click Proceed, instead of stalling on the choice.
Intent Signals, Captured
Every conversation is captured as a downstream signal, so Ring’s product and CX teams see exactly where borrowers hesitate and why.
Pilot Snapshot
By the Numbers
Four months live on the Available Offers page, measured across the full conversation set.
3,426
Total conversations started
76%
Overall proceed rate (latest week)
1.3
Conversations per user
18%
Deep engagement, over 120 seconds
0.4%
Asked for a human agent
20%+
Drop-off reduction already achieved
904
Calls in the March deep-dive sample
66s
Average call duration
The proceed rate has climbed week over week. The earliest March sample of 904 calls sat near 60%, the mid-pilot figure was about 71%, and the most recent week reached 76%. Latest figures are used as the headline.
Volume & Engagement
How Long Borrowers Stayed
From the March deep-dive of 904 calls. Most conversations were short and purposeful, with a meaningful share going deep on the numbers.
Top User Queries
What Borrowers Asked Riya
The four topics that drove the most conversations, and the product gaps the agent surfaced for Ring's product and CX teams.
Voice of the Customer
Beyond Conversion: What Borrowers Told Us
Moving borrowers to Proceed was the headline. The quieter win was the data. Every conversation was an unfiltered read on what borrowers care about at the offer stage.
47%
Perceived inflexibility
Borrowers wanted a higher or lower amount than the system offered, and 21% were openly frustrated they could not modify terms. Someone asking for ₹11,000 who is shown ₹17,000 wants to know why. A clear signal to make the amount feel explained, not fixed.
“क्या ग्यारह हज़ार का loan नहीं मिल सकता?” (Can I not get an 11,000 loan?)
29%
Multiple lenders, cognitive load
Showing three or more lenders with different EMIs needed explaining. Borrowers struggled to compare offers without a clear side-by-side view, so Riya walks them through it.
25.6%
explicit · 68% implicitEMI and rate transparency
A quarter asked about EMI directly, and far more implied it. This is financial literacy in action, borrowers doing total cost-of-borrowing math before committing.
“ईएमआई कितनी होगी?” (What will the EMI be?)
28.2%
explicit · 92% implicitApproval uncertainty
Borrowers wanted to know their odds before committing. Surfacing an approval probability up front closes the trust gap between a preliminary offer and a real one.
“क्या approval मिल जाएगा?” (Will I get approval?)
28.6%
Lender identity
Borrowers mentioned specific lenders, but it was rarely the deciding factor. The takeaway: rate and terms matter far more than any single brand.
7 calls
Tenure flexibility
A smaller set asked about term options, which feed directly into the EMI. Secondary to amount and rate, but offering two or three tenure choices still shapes the final decision.
Every one of these is fixable through script tuning and product changes, which means each week makes Riya sharper and the page smoother. Ring did not just get a conversion tool. It got a live feed of borrower intent flowing straight into product, pricing, and CX decisions.
Why Users Drop
Where the Real Friction Sits
The biggest drops are product gaps, not agent gaps. Riya is surfacing them, and holding the conversation while she does.
1 in 2
conversations show a borrower who is not happy with the default offer (41% dissatisfaction).
6 in 10
conversations ran longer than 30 seconds, with Riya able to help on the borrower’s query.
0.4%
asked for a human. The borrowers who stayed trusted Riya and her answers.
Even with the agent live, around 29% of borrowers who chatted on Offers still dropped later at the Final Offer screen, the gap the next phase is built to close.
Before & After
Before RevRag AI
After RevRag AI
Not just a chatbot. A conversion layer built specifically for the Ring lending journey.
What's Next
Covering the Full Lending Funnel
The Offers page is one step. The next phase extends Riya across every screen so no drop-off goes unaddressed.
Customise Loan Offer
- 41% are unsatisfied with the default offer
- Riya explains amount and tenure, then nudges to proceed
- Expected: a clear drop in Customise-page exits
Available Offers Page
- Live 4 months, proven 76% proceed rate
- Explains EMIs, lenders, and fees in plain language
- Nudges users to select multiple lenders before proceeding
Final Offer & Accept
- Users get confused by disbursal vs loan amount and next steps
- Riya explains processing fee and repayment schedule
- Expected: a 5 to 10% conversion uptick
Drop Chase
- Any drop at any stage gets reached out to
- The call uses full in-app conversation context
- Personalised script: the exact offer, page, and hesitation point
Customise → Offers → Final Offer → Drop Chase. Every drop-off in the funnel covered.
The Vision
Ring's Full AI Lending Companion
Today: Available Offers Page
Agent live, 76% proceed rate, four months of proven data, and a 20%+ drop-off reduction already achieved.
Next 60 days: Customise + Final Offer
Riya on the Customise Loan Offer and Final Offer pages, addressing the 41% dissatisfaction gap and recovering users who dropped.
Next 30 days: Drop Chase at scale
AI calling for every drop across all three stages, with context-driven personalised scripts. A closed loop, no warm lead left behind.
The vision: Full Lending Companion
The agent as a persistent guide across the entire journey, in-app and off-app, with intent data feeding offer personalisation. Ring’s AI relationship manager.
The goal
Every Ring user gets a personal loan guide, from first offer to successful disbursal, and back again if they ever drop.
Proven Across Banking & Fintech
RevRag AI's Track Record
Ring is one of several in-app agent deployments delivering measurable conversion lift across Indian banking and fintech.
PhonePe
12%
conversion uptick
AI agent lifting insurance renewal rates and activating users by guiding them on next steps in the app.
Motilal Oswal
33%
conversion uptick
In-app AI agent on the payments page, resolving payment and plan-related queries in real time.
Vivamoney
8.2%
conversion uptick
AI agent qualifying personal loan leads and helping them through the next steps.
Conclusion
For Ring, the Offers page was always make-or-break. Borrowers arrived ready but left with questions. RevRag AI's in-app agent closed that gap, answering every question instantly, in plain language, and moving more borrowers to Proceed.
The result was a proven, stable conversion lift at the exact point where loans are won or lost, plus a continuous read on what borrowers need next. Same page, far fewer people walking away from it, and a clear path to cover every other step in the funnel.