The Product
A Bangalore-based fintech startup was building a cash flow management platform for small and medium businesses. The core idea: SMBs in India have terrible visibility into their cash position because their financial data is scattered across Tally, bank statements, pending invoices, and their accountant's head. The startup wanted to change that by aggregating this data and giving business owners a clear, real-time view of their money.
They came to us after a previous development attempt — six months of work with a freelancer team — had produced a codebase that the founding team (both non-technical) had been told was "almost ready" for three months running. We reviewed the existing code and gave them an honest assessment: the architecture had fundamental problems that would be expensive to fix. We recommended starting over with proper architecture, and they agreed.
The Architecture Challenge
Cash flow management for SMBs has specific technical requirements that make it harder than it looks:
Multi-tenancy with strict data isolation. Each business customer's financial data absolutely cannot leak to other customers. This is not a theoretical concern — a data leak in a financial product is a company-ending event. We used PostgreSQL with row-level security policies enforced at the database layer, not just the application layer. Even if application code had a bug, the database would refuse cross-tenant data access.
Bank integration. Reading bank account data required integration with account aggregator frameworks (India's AA ecosystem) and direct bank statement parsing for banks not yet on the AA network. The bank statement parser — handling PDF statements from 40+ Indian banks with different formats — took two weeks of dedicated engineering time.
Real-time data freshness. Business owners checking their cash position want current data, not data from last night's batch job. We built a webhook-driven system where bank transactions trigger immediate updates to the cash flow model, rather than periodic polling.
What We Built (Over 7 Months)
The platform has three distinct parts:
The business owner web application: The core product — a Next.js dashboard showing cash position, upcoming payables and receivables, cash flow projections, and alerts for potential shortfalls. Designed to be usable by non-accountants; the target user is a business owner who understands money but not accounting software.
The data ingestion layer: Node.js services handling bank account connections, transaction normalisation, and categorisation. This processes raw financial data and transforms it into the structured model the dashboard queries. Runs continuously in the background.
The accountant portal: Many SMBs have accountants who manage their books. We built a separate access mode for accountants — they can see all their client accounts in one view, make adjustments, and handle the reconciliation work that business owners shouldn't need to do themselves.
Subscription Billing
The startup's pricing model had three tiers — starter (up to 2 bank accounts), growth (up to 10 accounts), and enterprise (unlimited, with white-labelling for accountants). We implemented this with Razorpay Subscriptions for recurring billing, with proper handling of plan upgrades mid-cycle (prorated charges), failed payment dunning, and automatic feature gating based on subscription tier.
The billing system took two weeks to implement properly — mostly because edge cases (what happens when a payment fails on upgrade day? What happens when a customer downgrades and has more accounts than the new tier allows?) required careful thought and specific user flows.
Compliance Considerations
Financial data in India sits under RBI guidelines and various data localisation requirements. All customer data is stored on AWS Mumbai (ap-south-1). The privacy policy and data processing agreements were reviewed with a legal advisor before launch. Encryption at rest and in transit, audit logs for all data access, and a clear data deletion process were all built in from the start.
Results
The platform launched 7 months after we started. In the first three months of operation, the startup onboarded 340 paying SMB customers, raised a pre-Series A round from a Delhi-based VC firm, and was featured in two fintech publications for its clean design and the quality of its bank statement parsing (the part that most competitors handle poorly).
If you're building a fintech product or a B2B SaaS platform with compliance requirements, reach out. This is exactly the kind of work we're set up for.