From registers and WhatsApp groups to one connected system
An FRP composites manufacturer in Gujarat — moulded gratings, structural profiles,
handrails — run for years on paper registers, Excel and three WhatsApp groups. This is what we built,
in what order, and what actually changed.
The business wasn't broken — it was running the way most mid-size manufacturers run.
Orders lived in a diary and in the owner's head. Costing lived in a spreadsheet with a filename like
costing_2026_new(3).xlsx that only one person really understood. Production status lived in
WhatsApp groups, which meant the answer to "Line B kitna hua?" depended on who replied and when.
The stock register was updated "mostly". And the month-end MIS took late nights to assemble —
after which the plant would dispute half its numbers anyway.
None of that fails loudly on any given day. It fails quietly, everywhere: a dispatch
promised on stock that wasn't there, a price quoted below cost because the sheet was stale,
a payment chased three weeks late because nobody owned the follow-up.
The problem was never the people. It was that every department kept its own
version of the truth, and the versions disagreed.
What we built — in the order it paid for itself
No big-bang ERP rollout. One module at a time, each one live and earning its keep before
the next began. The first went where the bleeding was worst: knowing what was ordered and
what it actually cost.
Order registerEvery order with its status, promised date and history — replacing the diary and the owner's memory.
Costing & accountsJob costing from actual consumption; invoices and payment follow-up in the same chain.
Production boardJobs, line status and output against plan — the WhatsApp group's replacement.
Stores & stockLive balances with reorder points that actually fire, and a movement ledger with references.
QC logsChecks and rejection reasons captured at the table, on a phone — not loose sheets.
MaintenanceSchedules and history per machine, so breakdowns stopped being a surprise twice.
Everything as web apps — one link that works on
the shop-floor phone, the stores tablet and the office desktop. Nothing to install, nothing to update,
and the tools keep working offline when the plant internet doesn't.
The screens
These are from our demo builds — the same design system and module logic we use in real deployments, seeded with neutral sample data, open for anyone to try in the
live demos.
The management view — revenue against target, margin, plant load, and the exceptions that need a decision today.
Stores — live balances, days of cover at real consumption, and a reorder list with one-click purchase orders.
The price book — every SKU's cost, price and margin, with anything under the floor flagged for repricing.
What changed
The month-end MIS became a button, not three late
nights — and the plant stopped disputing it, because the numbers came from the plant's own entries.
Stock stopped surprising anyone. Reorder points
fire before material runs out, and every movement carries a job or order reference, so variances
get caught in days rather than at year-end.
Quoting got safer. Costing comes from actual
consumption, so the margin on a quote is known before it goes out — not discovered after dispatch.
The 11:42 pm WhatsApp question disappeared.
Line status is on a screen anyone can open, which turned out to be most of what the groups were for.
Follow-up runs itself. Receivables ageing is
visible daily, with reminders — instead of fourteen awkward phone calls a week.
An honest note. This work was built inside our founder's own FRP
manufacturing business — which is exactly why it works the way a plant actually runs. We publish it
as our reference implementation rather than dressing it up as an external client win. Your deployment
starts from your item master, your process and your people, and the first module is live in
four to eight weeks.
