The problem
A repair counter is a busy, messy place. A customer hands over a phone with a cracked screen and a story. Someone writes it in a register, the phone goes in a drawer with a slip of paper, and from then on the questions start. Is it ready? Which technician has it? Did we use the new display or the old stock? Was this the same phone that came in last month with the same fault?
The repair business we built this for had outgrown that. They were also running billing in one place, stock in another and customer messages from someone's personal WhatsApp. The first version we made for them, back in July 2025, was a plain PHP app that did the basics. It worked, which is exactly why it kept growing.
In 2026 we started asking a different question: could other shops use this too? Our plan for it was clear about what not to build: a focused product for repair businesses, not another general billing app. It had to connect the customer, the device's IMEI, the job card, the technician, the parts, the bill, the warranty and the WhatsApp message in one place.
What we built
A rewrite that could run beside the old app
We rebuilt it in Laravel, but on the same database tables as the old PHP version. That meant both could run side by side during the switch-over, with no weekend of data migration and no day where the counter stopped. Old invoice links customers already had on their phones still work: they redirect to the new address.
A small core, and everything else as modules
Every shop is different. One sells used phones, another does B2B repairs on credit, a third wants staff attendance. So the core is kept small (customers, devices, invoices, products, users) and everything else is one of 26 modules that can be switched on or off under System, Plugins. Job tracking, GST, inventory, warranty, WhatsApp, quotations, the customer portal, payments and reports are all modules.
GST done the way an accountant expects
Invoices pick CGST and SGST or IGST from the place of supply and print the tax summary and the amount in words. Each branch can have its own GSTIN and its own invoice series. One rule took real thought: an invoice can only be deleted if it is the latest number. Anything older is cancelled instead, and still reported in GSTR-1 as cancelled. There are no gaps in the numbering for anyone to explain later.
WhatsApp without forcing a monthly bill
There are two ways to message a customer. Click-to-chat costs nothing: staff tap a button and WhatsApp opens with the message written. Or the shop connects Meta's WhatsApp Business API and messages go out by themselves: received, ready for pickup, a reminder to collect, and a nudge at 15, 30 and 60 days for phones nobody has come back for. Every repair shop has a drawer of those.
The things you only learn at the counter
Many features came from watching the shop work rather than from a list. An intake checklist with photos, so nobody argues later about a scratch. IMEI history, so a comeback repair is spotted. A stolen-phone check. QR stickers for the racks. Good, better and best repair quotes the customer approves online. Technician incentives. None of these are clever on their own. Together they are why the shop runs every repair through it.
Where it stands
The software is in daily use at the repair business it was built for, and it's backed by 86 automated test files across the core and the modules. On the website, SaralFix now has its own pages, a single simple price per staff login with every feature included, and a free demo booking.
It's honest to say where it stands. SaralFix is in early access. Today each shop runs on its own installation, and before we open sign-ups to everyone there is real work left: keeping every shop's data fully separate in one system, retiring the old app's password format, a proper data export for any shop that leaves, per-shop branding for the white-label plan, and online sign-up with billing. Early-access shops get the setup and their data transfer done by us, which is also how we learn what the next shop needs.
The lesson we took from it is the reason this case study exists. The best product ideas we have didn't start as products. They started as one client's problem, solved properly and used every day.
