All work
Finance

Aurora Banking

A multi-tenant digital banking platform with real-time ledgers and AI insights.

Client
Aurora Financial
Year
2025
Duration
7 months
Our role
Strategy · Design · Engineering
Aurora Financial · 2025
Overview

The project

Aurora needed to launch a modern, compliant digital banking platform serving multiple business tenants — with real-time balances, payments and AI-driven financial insight.

The challenge

Legacy core systems, strict compliance, and a need for real-time accuracy across thousands of accounts. Previous vendors couldn't deliver the performance or the multi-tenant isolation required.

Our solution

We built a multi-tenant platform with atomic ledger math, event-driven balance recomputation, role-based access, and an AI insight layer that explains spending and forecasts cash — all on a secure, audited foundation.

Services

SaaS DevelopmentAI SolutionsWeb Applications

Technologies

Next.jsNestJSPostgreSQLAWSStripeOpenAI
Technology

The stack behind it

Frontend

Next.js · React · TypeScript · Tailwind

Backend

NestJS · Node.js

Database

PostgreSQL · Redis · Prisma

Cloud

AWS (ECS, RDS, S3)

DevOps

Docker · GitHub Actions · Terraform

AI Integration

OpenAI · RAG · evaluation harness

Third-party APIs

Stripe · Plaid · KYC provider

Security

RBAC · encryption at rest · audit logs

Architecture

Multi-tenant, event-driven microservices

Case study

How it came together

01

Business problem

Aurora's legacy core couldn't support multi-tenant, real-time digital banking — and compliance risk was high.

02

Research

We audited the core systems, mapped compliance requirements, and interviewed ops teams to find the real bottlenecks.

03

Strategy

Build new alongside legacy: a modern, tenant-isolated platform that integrates with the core and migrates incrementally.

04

Design process

A calm, trustworthy product design and a reusable design system for web and partner portals.

05

Development

Atomic ledger math in transactions, event-driven recomputation, RBAC, and an isolated AI insight service.

06

Testing

Automated tests, load testing to thousands of accounts, and a full security review before launch.

07

Deployment

Zero-downtime rollout on AWS with IaC, observability and staged tenant onboarding.

08

Results

On-time launch, enterprise tenants live in weeks, sub-second dashboards, zero audit findings.

09

Lessons learned

Isolating the AI service from money movement (read-only) was key to both trust and compliance.

UI/UX highlights
  • A calm, trustworthy dashboard with real-time balances
  • Accessible, WCAG-AA flows for sensitive financial actions
  • A design system reused across web and partner portals
Architecture

Tenant-isolated services behind an API gateway, an event bus for ledger updates, Redis-cached aggregates for instant dashboards, and an isolated AI service that only ever reads computed figures.

Business impact

Aurora launched on time, onboarded its first enterprise tenants within weeks, and now runs thousands of accounts with sub-second dashboards and a clean compliance posture.

99.99%

Uptime

<300ms

Dashboard load

7 mo

Idea to launch

0

Security findings at audit

Gallery

Inside the product

Dashboard overview
Real-time ledger
AI insights panel
Mobile balances
Surface IT shipped what two previous agencies couldn't — on time, and built to scale.
Imran RahmanCTO, Aurora Financial

Have a project like Aurora Banking?

Let's talk — a senior engineer replies within one business day.