I am building the O9X ERP to manage 1,000+ associates in a low-margin, compliance-heavy environment. In my business, moving fast without a mapping phase is a liability. A single automated logic error in payroll does not just create a bug — it creates a systemic legal violation that compounds every month.
To reduce this risk, I undertook a detailed log-4-mapping-operations-before-codingoperations mapping phase before writing code. This log documents two specific compliance problems caught during that phase.
The Systems Logic: Why We Map
The primary goal was to catch legal edge cases where customer contracts deviate from Indian labor law. If the ERP replicates a contract clause that is legally invalid, we as the operator remain liable for the shortfall.
I brought in a labor compliance consultant to pressure-test my O9X payroll mapping. The question was not “is the code clean?” The question was “is this formula legal?”
Build Snapshot: The ESIC Validator
Edge Case 1: The Variable Overtime Under-contribution
The situation: all of our contracts assume a pass-through logic — if a customer does not pay for a specific compliance item, we do not account for it in payroll calculations. One or two customers per month exclude ESIC from overtime wages.
The mapping discovery: O9X’s draft formula excluded variable overtime from ESIC calculations.
The legal gap: under the Employees’ State Insurance Act, contributions of 4% of wages — 3.25% from the employer and 0.75% from the employee — must be calculated on all wages, including variable overtime.
The direct hit: for 10 employees who worked variable overtime in Q4 2025, this was a direct under-contribution.
The fix: O9X now enforces a law-first calculation. The pricing tool was updated so we explicitly communicate that ESIC applies to overtime wages and gets billed on actuals. The same change was made in the payroll module, where such charges are now included in ESIC deductions.
Edge Case 2: The Eligibility Mirage (Base vs Gross)
The situation: some contracts define ESIC eligibility based on gross pay at Rs 22,000.
The mapping discovery: because Rs 22,000 is above the Rs 21,000 legal threshold, those customers excluded ESIC for those employees.
The technical error: eligibility is determined by base pay, not gross. If the base is Rs 18,000 — below the threshold — the employee is eligible, even if overtime takes their total take-home to Rs 30,000.
The fix: 2 out of 18 contracts had this mismatch. We moved those guards to the correct plan. Since those contracts renew in April 2026, we could not trigger a mid-term customer audit. We reduced the guards’ additional pay by 3.25% to fund the employer contribution. The guards accepted this because they gained state medical benefits they had not previously had.
The Counterfactual
Without the 11-week mapping phase, O9X would have shipped with correct code executing incorrect logic.
The ESIC overtime error would have been automated into every payroll run and surfaced only during a government audit, carrying significant fines. The contract mismatch would have replicated an illegal clause and gone undetected until the April 2026 renewal — too late to correct without disruption.
I found multiple such edge cases during the mapping phase. Each is minor in isolation. Together, they add up to penalty exposure or contract loss.
Operator Invariants
Two rules are now fixed for every module I build. First: the law is the primary key; the customer contract is secondary. When they conflict, the system must flag a compliance variance. Second: no module is built until the worst-case audit is documented.
What I Am Still Figuring Out
I need to run a manual calculation on Q4 2025 to quantify the exact Rs amount of liability across all issues identified. A full audit of the remaining contracts is also required before their renewals. And while the planning phase clearly prevented rebuild cycles, I have not tracked the exact time saved against the longer planning period.
Planning caught compliance problems before they became automated violations. Fast iteration works for consumer applications. In compliance-heavy operations, it automates mistakes faster than you can catch them.