Railindia - IRCTC redesigned with accessibility

A platform fifty million people use, that fails its users in measurable ways.

A national utility, broken in measurable ways.

What worked, what broke, what I changed.

Automated tools catch about 30% of WCAG issues. The other 70% need a human, a keyboard, and a screen reader.

Five patterns across the 47 findings.

Findings on their own are a checklist. Insights are what those findings mean — the design decisions that produced them, the patterns that recur.

How IRCTC compares to its private alternatives.

IRCTC has the inventory monopoly — only it can issue tickets — but several private apps re-skin the same flow with better UX. They're imperfect, but they establish a baseline of what India's rail booking experience could look like. The accessibility scores below are mine, on a 5-point scale, against four core WCAG criteria.

Three users the system fails today.

Personas are grounded in the demographic data above and in disability-organisation interview research (Score Foundation, NAB, Mitra Jyothi published profiles). They represent the three most common ways IRCTC fails: low vision, CVD, and full blindness with a screen reader.

Asha books a train, before and after.

Asha — the 64-year-old retired teacher with mild cataracts — tries to book a Lucknow → Delhi ticket. Six steps. Where the system fails her today, and where it doesn't in the redesign.

Five rules, one design system.

Each decision is derived from one or more findings and informs every screen that follows. They double as the acceptance criteria for the design system.

What was tested, what comes next.

The audit is heuristic-led and assistive-tech-validated. The redesign is right, but it isn't yet proven. Honest framing of what this work has and hasn't yet earned.

What changes. What doesn't.

Every feature on IRCTC's booking flow is preserved. The change is in how each one is communicated — and in adding the patterns that aren't there at all.