The view from behind the counter

Almost every civic prototype is built for the citizen and forgets that there is a person on the other side of the window. The department is the largest single victim of this problem.

The arithmetic

1.74 crore rejections a year is 1.74 crore files opened twice. Examined, rejected, closed, and then re-opened when the citizen files again with the same defect. If pre-submission validation converted even one in five of those into a first-time-right claim, that is roughly 35 lakh fewer file re-openings a year across 135 field offices.

An estimate from public aggregates, not a measured result.

What an officer would see

• An incoming claim already pre-validated against all 61 checks, with the full trace visible.
• The conflicting records attached side by side, with the invoked clause quoted.
• One decision to make, instead of an investigation to run.

Three ways a department could adopt this, easiest first

• Lowest friction: take the open rule bundle and run it as a validation library in front of your own claim intake. No new portal, no new citizen habit, no new UI.
• Middle: give the assisted console to common service centres and to the outreach camps that already run monthly.
• Highest: the citizen-facing app. We would start at the bottom of this list, not the top.

Why this is not a criticism of EPFO

EPFO 3.0 is a real reform and it worked: auto-settlement, a three-day target, and 98.7% of claims inside 20 days. Throughput is solved. What a faster pipe cannot fix is a record that disagrees with Aadhaar. and that is the entire residual. Kosh is the pre-flight layer that lets that reform reach the people it currently cannot serve.

This page is a written argument, not a running console. The console is specified and not yet built; see what's real.