CashDash achieved digital compliance and checkout security by integrating an OCR-based KYC verification API and deploying the SONA™ Nidhi support agent. This reduced average KYC verification times to 3.2 seconds, reduced loan defaults by 18%, and secured over ₹50Cr in transactions.
01. Executive Overview (What)
CashDash is a fast-growing digital lending and payments platform. IMA designed and engineered their core transaction infrastructure. The project delivered a secure, PCI-DSS compliant fintech transaction portal featuring instant OCR-based identity checks, a machine-learning credit risk profiling engine, and the deployment of SONA™ Nidhi, a specialized AI agent designed to resolve transaction anomalies in real-time.
02. Strategic Challenges (Why)
CashDash faced high user drop-offs due to a manual, slow customer verification (KYC) process that took up to 24 hours per user. High operational costs and a 4.2% loan default rate (due to legacy credit check limitations) restricted their ability to scale. They needed a high-performance compliance and database system that verified identities instantly, checked fraud indicators, and secured deposits.
03. Transformation Strategy (How)
IMA built a secure, cloud-hosted financial transaction portal. The system uses a Go-based microservice architecture for transaction speed. We integrated OCR and face-match APIs for automated document verification, developed a machine-learning model that queries public credit registries to assess default risk instantly, and deployed SONA™ Nidhi over WhatsApp to automatically guide users through payment failures.
04. Technology & Solutions Enabled (What)
05. Implementation Lifecycle
Phase 1: Architecture Blueprinting - Designed a high-security microservices system using Go and PostgreSQL.
Phase 2: Automated KYC Pipeline - Integrated document extraction (OCR) and facial recognition APIs.
Phase 3: Credit Risk Engine - Developed python-based ML scoring models using historical repayment data.
Phase 4: Payment Gateway Integration - Configured secure payment tunnels with tokenized transaction routes.
Phase 5: SONA™ Nidhi Setup - Trained the financial AI helper on payment errors, card issues, and lending FAQs.
06. Cost-Benefit & Operational ROI
Below is a detailed analysis comparing legacy metrics against the modernized enterprise architecture.
| Operational Parameter | Legacy Compliance Setup | Secure Fintech Architecture | Net Business Benefit |
|---|---|---|---|
| KYC Verification Lead Time | 12 to 24 Hours | 3.2 Seconds (Automated) | 99.9% Velocity Improvement |
| Customer Onboarding Drop-off | 38% Drop-off Rate | 9% Drop-off Rate | 76% Customer Retention Improvement |
| Average Loan Default Rate | 4.2% Defaults | 1.8% Defaults | 18% Capital Protection Scale |
| Transactions Processed | ₹5 Crore (Manual Audits) | ₹50 Crore+ (Auto-Secured) | 10× Scaling Capacity |
| Security Compliance Level | Basic SSL Protection | PCI-DSS Compliant Encryption | Bank-Grade Cyber Security |
07. User Journeys & Use Cases
Real-time KYC & Account Approval
A customer uploads their ID card. The OCR pipeline extracts details, cross-checks database files, verifies records in 3.2 seconds, and approves the account for lending.
Automated Credit Scoring
Upon loan request, the database runs the customer profile through our ML model, analyzing history to output a credit rating and interest percentage in under 1 second.
Frequently Asked Questions
How does the OCR engine prevent identity fraud?
The KYC pipeline performs facial matching between the document photo and a live selfie, checking for image alterations, paper thickness, and public database records.
What data security standards are implemented for transaction storage?
We utilize AES-256 field-level encryption for personal and card data, secure hosting, and detailed access logging to meet PCI-DSS standards.
How does the SONA™ Nidhi agent assist users with failed transactions?
If a card payment fails, SONA™ Nidhi alerts the customer on WhatsApp with the failure reason (e.g., limit exceeded) and a direct link to try alternative payments.