Six months ago, a CTO at a mid-sized NBFC asked us a simple question: “Why does every new RFP we get mentions React and Vite specifically?” He wasn’t wrong to notice. Over the last two years, we’ve seen this pairing go from “the new hotness” to the stack that procurement teams write into requirements documents before a single line of code is discussed.
At Speqto Technologies, we’ve migrated three BFSI clients off Angular.js and Create React App in the past 18 months, and the pattern is consistent enough that it’s worth writing about honestly – including where this stack isn’t magic.
The problem nobody wants to admit out loud
Most enterprise web apps in lending, insurance, and payments were built between 2016 and 2020, on Angular.js, older Angular versions, or Create React App. These weren’t bad decisions at the time. But three things changed:
- Angular.js hit end-of-life in January 2022, and banks running loan origination systems on it are now carrying real security risk with no vendor patches coming.
- CRA’s build times became untenable as codebases grew. One of our fintech clients, a payments aggregator processing UPI and card transactions, had a dev server that took 47 seconds to reflect a one-line CSS change. Developers were losing 20-30 minutes a day just waiting.
- Compliance and audit requirements demand faster iteration on KYC flows, RBI-mandated disclosure screens, and consent management UIs. Slow tooling directly slows down regulatory response time.
What Vite actually changes (not marketing, just mechanics)
Vite doesn’t bundle your app during development. It serves modules over native ES modules directly to the browser and only transforms what’s requested. That’s why cold starts that took CRA 15-20 seconds on a mid-sized codebase take Vite under 2 seconds.
For the payments aggregator client mentioned above, we measured this after migration:
- Dev server cold start: 18.4s (CRA) to 1.1s (Vite)
- HMR update on a single component change: 3-6s to under 200ms
- Production build time for their admin dashboard (140+ routes): 6 minutes to 1 minute 40 seconds
None of this changes what the end user sees. But it changes how fast a team of 15 developers can ship a fix when a regulator flags an issue on a disclosure screen with a 48-hour remediation window. That’s not a developer-experience nicety in BFSI – it’s operational risk reduction.
Why React specifically, and not Vue or Angular
This is where enterprise buyers get it right more often than tech blogs give them credit for. It’s not about React being “better.” It’s about three boring, practical things:
- Hiring pipeline. In Indian tech hubs – Bangalore, Pune, Hyderabad, Gurgaon – React developers outnumber Vue and Angular developers by a wide margin. For a bank building an internal platform team, that matters more than framework elegance.
- Component libraries built for compliance UIs. Libraries like React Hook Form with Zod validation, or React Aria for accessibility, map directly onto RBI’s accessibility guidelines and the kind of multi-step KYC forms that BFSI apps are full of.
- Micro-frontend maturity. Large banks rarely have one team owning the whole app. Module federation patterns with React are simply more battle-tested for splitting a core banking portal into independently deployable pieces owned by different vendor teams.
A real migration, and what actually broke
For a housing finance client we worked with last year, the loan origination system was on CRA with Redux and Webpack 4. The migration to Vite wasn’t a weekend job – it took our team about three weeks for a codebase with 280 components.
What actually caused friction wasn’t Vite itself – it was legacy CommonJS packages that assumed a Webpack environment, particularly some older PDF-generation and e-signature SDK wrappers used for loan agreements. We had to patch or replace four packages. Anyone selling this migration as a drop-in weekend project isn’t being straight with you.
What we didn’t have to touch: their existing component logic, their state management, their test suite written in React Testing Library. That reuse is the real win – you’re changing the build tool, not rewriting the application.
Where this stack isn’t the right call
If your BFSI client still has a meaningful user base on IE11 or very old Android WebViews used in point-of-sale devices at branches, Vite’s ESM-first approach needs extra legacy plugin configuration, and it’s worth testing thoroughly before committing. We’ve seen one insurance client where a legacy agent-facing tablet app forced us to keep a Webpack build alongside Vite for a transition period.
The honest takeaway
React and Vite are becoming default not because they’re trendy, but because they reduce two things enterprise buyers actually care about: time-to-fix on compliance-critical screens, and the cost of finding developers who can maintain the app three years from now. If your platform team is still debating this, the conversation is less about technology preference and more about how much operational risk you’re willing to carry on an unsupported or slow toolchain.
If you’re evaluating a similar migration for a lending, insurance, or payments platform, Speqto Technologies has done this migration path enough times to give you real numbers before you commit budget – not just a sales pitch.
