AM Ayush Mutha
THE AUDIT

THE WORK

A practice took thirty-eight people off a shared Excel timesheet.

One build, described properly, rather than a wall of logos. It is here because it is the one carrying real users today — and a system in daily use is a different claim from a demo that runs. Every figure below came from counting.

CA FIRM · MUMBAI

IN PRODUCTION

Staff time and job costing, off the spreadsheet

The real defect was not the file, it was the data going into it. Their template carried a 500-item client dropdown and a second, separate 500-item client group dropdown with nothing linking the two — so the instructions asked every person to hand-check that the two agreed, on every row they typed. That is not a formatting problem. It is job costing built on data nobody can trust, and no amount of chasing timesheets fixes it.

Live
IN DAILY USE SINCE 25 AUG 2026
38 staff
ACROSS AUDIT, TAX AND GST
700+
AUTOMATED TESTS ON EVERY CHANGE
Their box
THE FIRM’S OWN CLOUD ACCOUNT

THE CLIENT LIST NEVER LEAVES THEIR SERVER · NO DATA PASSES THROUGH ME, AND NONE OF IT APPEARS ON THIS PAGE.

WHAT IT ACTUALLY DOES

  • Weekly time entry

    A week at a time, entered against a client and an assignment type, with the day’s hours totalled as you type and a cap that stops a day quietly exceeding itself.

  • Leave, end to end

    Apply, preview the working days a request will actually consume, approve or reject, withdraw, cancel. Sundays and holidays are handled explicitly rather than silently shortening a request.

  • The week locks itself

    Entry closes on a schedule instead of by email reminder, and an admin can reopen a period deliberately when someone genuinely needs it.

  • Managers see their own people

    A team view with drill-down for whoever a person reports to, and a firm-wide view for admins — not one shared file where everybody can read everybody.

  • Reports and Excel out

    A pivot by month and by person, drill-down to the underlying rows, and exports for both time and leave — so the spreadsheet is an output, not the system of record.

  • Master data with a spine

    Clients, client groups, employees, segments, assignment types, leave types, holidays and financial years are all first-class records — which is what kills the two-unlinked- dropdowns problem at its root.

NO CLIENT NAMED · NO CLIENT DATA ON THIS PAGE · FIGURES MEASURED AUGUST 2026.

WHY ONLY ONE

The short list is the point.

I have built more than this. The rest are not here yet, and the reason is the same in every case: either nobody has scored them against a hand-checked set, or they are not carrying real users. A tool that runs is not evidence that it is right — those are different claims, and only one of them is worth paying for. When a build earns its number, it gets a section here.

HOW IT STARTED

It did not begin as a build.

It began by watching the work happen and costing what it was losing. That is the audit: two weeks, every leak found and given a rupee number, and a fixed price against each fix. ₹10,000, credited in full against anything you then build.

WHATSAPP +91 87670 73780 READ What the ₹10,000 audit covers WATCH A document becoming a row