Six years leading design across fintech, EdTech and commodity trading at a publicly listed enterprise software company.

3i Infotech is a publicly listed enterprise software company (NSE: 3IINFOLTD) with a portfolio that spans BFSI, EdTech, governance & compliance, and commodity trading. Each product line had grown organically since the early 2000s, with different teams, different conventions, and different visual languages. A client buying two modules ended up with two incompatible UIs, two training programmes, and two sets of cosmetic complaints in every RFP.

When I joined in 2018, design sat downstream of engineering: dev shipped, design polished, and neither side owned the overall experience. Over six years my remit expanded from "redesign a screen" → "own UX for a product" → "own the cross-product system and mentor the team."

"Every client demo started with an apology — 'I know this product looks different from the last one we showed you, we're working on it.' We needed that apology to go away." — VP Products, first year kickoff

Core problems:

• Visual fragmentation across 5 products serving overlapping enterprise customers.
• Dense, legacy screens — built for keyboard power-users, impenetrable to new joiners.
• No research function. Feature decisions were made by sales + engineering, with no end-user input in the loop.
• Tall stakes for every change. Core banking can't ship a broken form. A trading platform can't ship a misleading price. A campus system can't ship an inaccessible exam screen.

In a 20-year-old codebase with 60+ engineers, a big-bang redesign is career suicide. I played a long game: pick small, highly-visible flows, ship measurable wins, and use those wins as political capital for bigger system work. By year 4, that capital paid for the cross-product design system and a research function that hadn't existed before.

• Year 1–2 — Quick wins inside flagship banking screens. The teller transaction flow was the trojan horse. Reduced clicks per transaction by ~30%. Demos got measurably easier.
• Year 2–3 — Brought research in. Field studies in 4 banks, contextual interviews with compliance officers, the first personas + journey maps the org had ever seen.
• Year 3–4 — Took on DMCC's commodity trading platform end-to-end. Then the NuRe Campus EdTech product. Then the NextGen GRC SaaS. Three different industries, one design language.
• Year 4–5 — Proposed and built the cross-product design system. Got buy-in by showing side-by-side before/afters in a client pitch.
• Year 5–6 — Mentored 4 designers. Represented design in RFP responses and client workshops. Stepped back from making every screen.

Of the five flagship products I led design on at 3i Infotech, three are public enough to talk about by name. They sit in different industries — commodity trading, higher education, governance & risk — but the design problems rhyme: dense data, regulated workflows, multiple stakeholder roles, and a need to make the right decision visible without dumbing the system down.

The system had to serve three contradictions at once. Density — risk officers and compliance teams scan hundreds of rows a day. Clarity — bank tellers and university students touch the product for the first time with minimal training. Trust — commodity traders and risk officers will not accept a system whose state they cannot independently verify. The solution was a dual-density grid, strict type hierarchy, a colour palette tuned for long-session screen reading on low-grade branch monitors, and an explicit "system state" pattern used everywhere automation was involved.

Core banking is the wrong place to ship experimental AI. Explainability, auditability, and regulatory reporting matter more than cleverness. So we picked surfaces where AI could be advisory, fully auditable, and reversible — never autonomous.

Illustrative numbers — I'll swap in verified figures before publishing.