Metisent

Launching a Full-Stack PPI and UPI Fintech Platform in 7 Months

Client

Global Consumer Finance Group

Industry

BFSI / Fintech / Digital Payments

Duration

Approximately 7 months

What This Enabled

A compliant PPI and UPI fintech platform that could reliably onboard users, process transactions, and generate behavioral data. This created a foundation for scalable customer acquisition and future lending products.

About the Project

Context and Strategy

Build a scalable PPI over UPI platform for digitally emerging users in Tier 2 and Tier 3 cities in India, delivering a mobile-first, low-friction experience. The platform needed to support Aadhaar-based eKYC onboarding, instant digital card issuance with optional physical card fulfilment, and seamless UPI transactions across payment flows.

It had to integrate hyper-local merchant rewards with real-time transaction processing, enabling cashback or reward point accumulation and redemption based on city-level ecosystems. Every transaction needed to be captured and structured to build financial behavior profiles that could later support credit underwriting.

The system was expected to handle onboarding and transaction volumes at market-entry scale, while integrating multiple external PSPs, card issuers, and payment gateways with differing APIs and failure modes. All of this had to be delivered within a fixed 7-month timeline, with full compliance to RBI guidelines for PPI, NPCI requirements for UPI, and reliable handling of asynchronous payment flows.

Understanding the Challenges

Regulated, Greenfield, and Time-Constrained

This was a greenfield build that had to operate correctly in a regulated, multi-vendor, asynchronous environment from day one. The complexity was not just in building features, but in ensuring that the system behaved predictably despite dependencies on external PSPs and payment infrastructure providers.

Fixed Deadline

A non-negotiable, public launch timeline of approximately 7 months. Every architectural decision carried a delivery cost and had to serve the deadline or it was the wrong decision.

Regulatory Complexity

RBI regulations for PPI, NPCI requirements for UPI, and Aadhaar-based KYC embedded into every flow. There were no shortcuts, and no feature could be deferred for compliance to be added later.

Vendor Fragmentation

Multiple external vendors including KYC providers, card issuers, UPI PSPs, and payment gateways with inconsistent APIs and failure modes.

Asynchronous Uncertainty

UPI and payment flows relied on non-deterministic callbacks, making consistency and reconciliation a core engineering challenge rather than an operational one.

Technical Gaps

No Flutter SDKs existed for any of the vendors. Every integration across KYC, card issuance, UPI rails, and rewards required custom native bridge work built from the ground up.

How We Approached It

Isolating Complexity, Enforcing Consistency

The primary problem was not building individual features, but making a multi-vendor PPI and UPI system behave predictably. The approach focused on isolating external complexity and enforcing consistency across all transaction flows.

Strategic Sourcing

Leveraged certified vendors for PPI issuance, UPI PSP integration, and KYC to reduce regulatory overhead and accelerate time-to-market. Building these capabilities in-house would have required independent NPCI and RBI compliance per feature.

Orchestration First

Built a central orchestration layer to standardise all vendor interactions and handle retries, failures, and callbacks. The mobile app never called a vendor API directly.

Decoupled Architecture

Ensured the application layer remained insulated from vendor instability and API variability. Vendor contracts, error models, and async behaviours were contained entirely in one layer.

Cross-Platform Efficiency

Used Flutter with native bridges to maintain a single codebase while supporting vendor SDK requirements. Where native was required for KYC and biometrics, custom method channels served as compliance bridges.

What We Built

A Production-Grade PPI and UPI Platform

A unified PPI and UPI platform that brought together multiple regulated capabilities while isolating vendor complexity behind a controlled backend.

End-to-End Fintech Capabilities

Aadhaar-based KYC across all levels, PPI card issuance and lifecycle management, and UPI payments covering collect, pay, and QR flows.

Orchestration Layer

Central backend managing all integrations with card issuers, PSPs, and gateways. All vendor surface area isolated behind one layer with no direct vendor calls from the app.

Shadow Ledger

Independent transaction ledger to ensure consistency across PPI and UPI flows and eliminate reconciliation gaps. Any discrepancy was caught before it reached the user.

Marketing BFF

Backend-for-frontend layer enabling dynamic offers and campaign updates without app releases or developer involvement. Campaigns could be updated post-launch without touching the codebase.

Impact

On Time, Compliant, and Stable from Day One

The platform did not just meet the launch deadline. It operated reliably under real-world payment conditions from day one, which is critical in PPI and UPI ecosystems.

On-Time Delivery

Zero to production in approximately 7 months, aligned with a fixed public launch. No compromise on scope or regulatory requirements.

Regulatory Integrity

Full compliance across RBI PPI guidelines, NPCI UPI flows, and Aadhaar KYC requirements from the first transaction.

Operational Stability

Handled high-volume onboarding and transaction loads without inconsistencies. The platform behaved reliably under real-world payment conditions from day one.

Transaction Correctness

Eliminated duplicate transactions and reconciliation gaps through idempotent processing and shadow ledger design.

Post-Launch Flexibility

Enabled continuous updates to offers and engagement without engineering dependency. The BFF-driven layer meant the marketing team could update campaigns without developer involvement.

Key Insights

In PPI and UPI systems, the hardest problems are not visible in the UI.

Systems fail at the boundaries between retries, callbacks, and reconciliation. That is where the architecture must be strongest. The visible surfaces: the app, the card, the flows, are what users interact with. The shadow ledger, idempotency keys, and event-driven async patterns are what make the platform trustworthy. In this domain, that invisible layer is the product.

Working on a fintech platform or a complex integration challenge?

Get in touch