Skip to content
Senthil Kumar
← All projects

Case study

Enterprise frontend platforms

Principal Developer

Architecture, standards, and squad delivery at scale

2024–Present · Enterprise · Architecture · Micro Frontends

Discuss this workRelated experience →

The choice

Rewrite the frontend — or split it so teams can ship?

When a large frontend gets slow and fragile, the tempting answer is a full rewrite. That usually stops delivery for months.

The easy option

  • Keep one giant app that every team ships into
  • Write standards in slides and hope people follow them
  • Plan a full rewrite when bugs pile up

What I chose

  • Split the frontend so teams can release on their own
  • Enforce standards in design and code reviews every week
  • Improve the current system first. Rewrite only a piece that is clearly separate

In short

The fastest rewrite is usually a careful split of the app you already have.

What I checked first

  • Can we change one piece at a time?

    If extracting a module freezes every other team, the plan is too big.

  • Can the business keep shipping?

    If work cannot pause, improve the current system in steps. Do not stop the world for a rewrite.

  • Do we understand the current app?

    Where the real behavior lives only in people’s heads, try a small slice first — then decide.

Related writing: When I refactor, and when I rewrite →

Context

Across banking and enterprise digital platforms, frontend systems must stay highly available, secure, and deliverable by multiple squads — without collapsing into a single fragile monolith.

This write-up stays intentionally high-level. No confidential systems, client data, or proprietary internals are discussed.

What I led

I steered frontend architecture and technical vision: aligning modern full-stack technologies with business goals, establishing architectural governance, and setting clean-code standards that reduced production defects and accelerated release cycles.

That included mentoring engineers across squads through design reviews, code reviews, and workshops — raising delivery quality while improving developer experience and reusable platform foundations.

Approach

  • Modular architecture patterns that cut technical debt and improved release velocity.
  • Micro frontend boundaries so teams can ship independently.
  • Standardized tooling for performance, testing, and predictable Agile sprints.
  • Ensured clear management of client-side state using Tanstack Query for scenarios requiring cross-team orchestration.

Tools on these platforms

  • React + TypeScript — type-safe UI at enterprise scale.
  • Micro Frontends — independent deployability across squads.
  • Tanstack Query — API client state management for cross-team orchestration.
  • Node.js, Express, REST — pragmatic API layers close to the frontend team’s mental model.
  • Agile / Scrum — structured delivery, reviews, and workshops as part of how the platform ships.

Impact

Reusable engineering foundations, stronger team standards, and platforms that are easier to extend — measured in fewer defects, faster releases, and healthier developer workflows.