Why we built our own production system

The most common question in garment manufacturing is four words long.

Where is my order?

Every manufacturer gets asked it, constantly, by every client. And in most of this industry the honest answer would be: somewhere in the building, probably fine, let me walk over and look.

What the client actually receives is a date. The date is frequently a guess with a confident tone. Then the date passes, and a new date arrives, and trust starts draining out of the relationship in a way that is very hard to reverse.

We built Saarvik because we had been on both ends of that conversation and did not want to keep having it.

What it looked like before

Not a system. A whiteboard, a lot of WhatsApp, and memory.

That combination genuinely works up to a point. When there are six orders on the floor, the person running production holds all six in their head, and they can answer any question about any of them instantly and correctly. It feels like plenty.

The trouble is the failure mode. It does not degrade gently — it works, and works, and then one week there are twenty-two orders instead of six and it stops working, without warning. Two orders are waiting on the same printing slot and nobody noticed until both were late. Someone remembers a stage was completed that was not. A client asks about an order that has been sitting behind another one for three days for no reason anybody can reconstruct.

Nothing dramatic goes wrong. You just start being slightly wrong about everything, and every wrong answer costs a little credibility.

What we tried first

Spreadsheets. Obviously. It is what everyone tries.

A shared sheet with a row per order and a column per stage, updated as things moved. On paper it is exactly the right shape for the problem.

It failed for a reason that had nothing to do with spreadsheets. The people who know a stage is finished are the people doing the work — Umesh at the cutting table, Raj at the press — and they are not going to stop, find a laptop, locate the right row and mark a cell. Nor should they. That is not their job and it is a bad use of thirty years of pattern-cutting experience.

So the sheet was updated later, in a batch, from memory, by someone who was not there. Which is the whiteboard again, with more steps and a false air of precision.

Then we tried off-the-shelf ERP software. That failed differently: it was built for a factory with departments, approval chains and procurement workflows. We are seventeen people in Ghatkopar. Configuring it would have taken longer than building something, and we would have spent the rest of our lives working around somebody else's assumptions.

What Saarvik actually does

Very little, deliberately.

Every order exists as a record. Every record moves through the stages an order actually passes through here — consultation, fabric, cutting, stitching, printing, quality, finishing, packing, dispatch. Each stage is marked off as it completes, by the person completing it, in a couple of taps on a phone.

That is the whole thing. It is not clever software. The design constraint was that marking a stage complete has to be faster than not bothering, because anything slower than that will not get done on a busy floor, and a tracking system nobody updates is worse than no tracking system — it is a confident, wrong answer.

Everything else Saarvik does falls out of that one honest record: which orders are where, what is queued behind what, which stage is becoming a bottleneck, and whether the date we gave someone three weeks ago still holds.

What changed on the floor

The thing we did not anticipate was the effect on us rather than on clients.

Bottlenecks became visible. Printing being the constraint on a given week is obvious in hindsight and was surprisingly hard to see in advance when the information lived in one person's head. Now it is on a screen, and you can move work before it becomes a late delivery rather than after.

The quieter change was that nobody has to remember any more. There is a real difference between a floor where the schedule lives in someone's head and one where it lives somewhere everyone can see. The first is stressful, and it concentrates risk in whoever that person is. If they are ill, you have lost the schedule.

What it means if you are the client

You can ask, and get an answer.

Not "it should be ready around the end of next week". An actual stage: your pieces are cut, they are in stitching, printing is next, dispatch is Thursday.

That is a low bar. It ought to be a completely unremarkable thing for a manufacturer to be able to tell you, and the fact that it is not is the whole reason this exists.

It also means when something does slip — because occasionally something does — you hear it from us early, with the actual reason, rather than discovering it on the day the delivery does not arrive. A supplier is late, or a stage is stuck. You would rather know on Tuesday than find out on Friday, and we would rather tell you.

Most of the trust in this business is not built by things going perfectly. It is built by what happens when they do not.