case stusy

Reimagining credit cards for digital engagement and cardholder empowerment

1.5

M+

App downloads

4.8

Average rating on Play Store & App Store

476

K

Active devices
Case Study
Industry :Fintech & Wealth Management
Platform :
Voice (Amazon Connect)  | Android | IOS
Services :Custom Software Development, UI/UX Design, Quality Engineering, Data & AI

BACKGROUND

A credit provider building for the omnichannel cardholder

Mercury Financial, now part of Atlanticus Holdings Corporation, is a US-based provider of consumer-friendly credit products. With a growing cardholder base, Mercury needed more than a well-designed app; it needed an end-to-end digital ecosystem serving cardholders across mobile, voice, and chat. Robosoft built that ecosystem through a native mobile app and an AI-powered Intelligent Virtual Assistant (IVA) for voice and chat.

THE CHALLENGES

Addressing

critical gaps

Digital control gap 

Digital control gap 

Mercury Financial's cardholders had no streamlined way to view balances, manage payments, or track transactions on mobile, pushing routine account queries to live agents, inflating support costs and creating a backlog that limited scalability.

Limited engagement

Limited engagement

Without personalized touchpoints, customers had little reason to engage beyond monthly payments, leading to missed opportunities in retention and credit-building.

Reliability 

Reliability 

With a growing user base across diverse devices and OS versions, ensuring consistent quality with every release was becoming increasingly difficult to manage manually.

Agent overload

Agent overload

Customer service agents were overloaded handling repetitive, low-complexity queries, leaving limited bandwidth for the complex cases that genuinely needed human judgment.

Integrated Approach

THE ROBOSOFT SOLUTION

An integrated approach

Native app development

Native app development

We built native Android and iOS apps using Kotlin and Swift respectively, with Android Studio and Xcode as IDEs, delivering platform-optimized performance and interactions.

Simplified card journeys 

Simplified card journeys 

We designed intuitive flows for balance views, payments, transaction tracking, and card management, giving cardholders instant, self-service control over their accounts. 

Personalization 

Personalization 

We integrated Braze for push notifications, reminders, and personalized offers across channels, driving timely engagement based on spending habits and payment behavior. 

Quality engineering

Quality engineering

TestNG framework-powered automated regression testing, with scripts covering critical happy flows, invoked via BrowserStack across multiple devices for cross-platform reliability.  

Conversational AI system

Conversational AI system

We rebuilt Mercury Financial's voice servicing channel from the ground up using Amazon Lex and Amazon Connect. The IVA handles the full range of account servicing interactions, such as balance checks, payment processing, card PIN updates, activation, and lost card reporting, through natural conversation in both English and Spanish.

Secure intelligent routing

Secure intelligent routing

Multi-step caller verification using ANI, card details, SSN, and date of birth, ensuring secure access at every interaction. Intelligent routing logic determines whether to continue assisting the caller autonomously or transfer them to a live agent based on account status, intent, eligibility, and compliance rules. 

Compliant architecture

Compliant architecture

Built-in compliance messaging, PD validation, and debt-collection logic aligned with regulatory requirements were embedded at the design stage. The system is audit-ready by default rather than patched for compliance later.

Voice-to-chat extensibility

Voice-to-chat extensibility

The IVA was architected from the outset to extend into chat-based channels. In Phase 3, a chatbot proof of concept was deployed on the Mercury website, replicating the same IVA capabilities over text, and a RAG-based architecture was enabled via Amazon Bedrock to answer general frequently-asked questions at scale.

Logging and monitoring

Logging and monitoring

CloudWatch and Splunk give the team structured logging and real-time monitoring across the IVA, supporting faster troubleshooting and continuous improvement as the system scales.

glow
KEY FEATURES
feature
VALUES DELIVERED

Clear outcomes

that drive growth

Mass consumer adoption 
Mass consumer adoption 

Mass consumer adoption 

The app crossed 1.5 million downloads and 476K active devices, establishing Mercury as a digitally accessible credit card brand in the US market. 

Top-tier user satisfaction 

Top-tier user satisfaction 

A 4.8 average rating across app stores reflected the quality of the cardholder experience, driving organic growth and lower acquisition costs. 

Cost reduction 

Cost reduction 

Transaction costs dropped from $4.25 to $0.10, transforming the unit economics of card servicing and unlocking significant operational savings. 

Voice servicing, rebuilt for the digital era

Voice servicing, rebuilt for the digital era

The IVA replaced a rigid, legacy IVR with a conversational, AI-driven servicing channel capable of handling end-to-end account interactions in both English and Spanish.

Improved observability

Improved observability

Structured logs and centralized monitoring across the IVA give the team clearer visibility into call flows and drop-offs, supporting faster development velocity going forward.

Build and scale financial solutions through secure engineering

RELATED WORKS

More work that

reflects our impact

View Allright arrow
Carousel Image
Seamless platform migration for advisors and investors
Paytm Case Study
Redefining payments and mobile commerce at scale
Carousel Image
A digital platform for distributor networks in asset management