Textiles7 min readNov 2026

The buyer asked at 9 pm. The rate card was on someone else’s phone.

Knitwear units live on WhatsApp for buyers and job-workers, while stock sits across godowns. Why that combination hurts in a tight export year, and how an enquiry-and-stock agent helps.

A merchandiser in Tiruppur will recognise this thread. A buyer — at home or overseas — messages at 9 pm: a rate for a style, a minimum order, a shade, a photo, “can you ship in 18 days?”. The person who knows the current rate is at a family function. The leftover stock from last season is in a godown that is not the one next to the office. A job-worker who has the fabric is being chased in a different WhatsApp group. By morning, the buyer has also written to two other units.

WhatsApp is not the problem. The cluster already runs on it — for buyers, for agents, and for dyeing, printing, embroidery and stitching job-workers who will not log into a heavy system at 10 pm. The problem is that the same chat is the catalogue, the production meeting, the stock register and the memory of the last order.

Life without an AI agent

Enquiries have no first owner. Rates, MOQs and pictures are copy-pasted when someone is free. A mixed Tamil-English or Hindi message waits for the one person who “handles that buyer”. There is no list of who asked, what was quoted, and who went silent after a sample went out.

Stock is split by definition. Greige, dyed fabric, trims and finished goods sit in more than one unit or godown. A “yes, we have it” is often a guess. A repeat order means reconstructing last season’s size break from a spreadsheet and a cutter’s memory.

Documents lag the goods. GST job-work challans, e-way bills and invoices are a separate scramble. Goods move; the paper follows. Payment reminders are personal favours.

This was always messy. It is more expensive when the cluster is already under pressure.

Why the leaks matter more now

Tamil Nadu textile exports fell to $7.1 billion in FY24 from $7.9 billion the previous year, according to government data reported by The Economic Times (2025). Industry voices in Coimbatore have also pointed to higher cotton costs, electricity costs and labour tightness. Industry bodies have said labour shortages can run at around 30% at any given time and worsen when orders surge, and wage bills have risen sharply in recent years (Mint, September 2024).

When margins are thin and people are short, you cannot staff a 24-hour merchandising desk. But the buyer’s clock did not change. They still write at 9 pm, and they still compare the first clear answer they get.

A Harvard Business Review study of online leads found that making contact within an hour was nearly seven times as likely to qualify the lead as waiting even an hour longer. A garment enquiry is not a software download, but the first unit that sends a correct rate, a photo and an honest lead time is the one still in the conversation at breakfast.

What an AI agent can take on — and what it should not

A practical first agent in this industry is not “AI for fashion design”. It is an assistant with your approved facts:

  • It replies to routine buyer questions, at night and on Sunday, in the languages you already sell in (English, Tamil and others you specify), with catalogue images, MOQ and lead time from your data.
  • It logs the enquiry and reminds a merchandiser if a sample or quote has had no follow-up.
  • It can be connected to stock by style, size, colour and godown, so “available” is not a guess, and a low-stock alert appears before the next cut.
  • When billing is connected, it can share a GST invoice and a payment reminder on the same WhatsApp thread.

It should not invent a price you did not approve, promise a ship date the floor has not confirmed, or pretend to be the owner. Special rates, quality complaints and new buyer negotiations stay human.

We also should not pretend this is a full mill ERP. Many Tiruppur units still combine Tally, sheets and WhatsApp, and that is the starting point. The first win is that a 9 pm question gets a truthful first answer, and leftover stock is visible before the next “yes”.

A published example

The Dreamwell homepage already carries this textile and wholesale example: Vellore Fabric Hub reduced manual workload by 92%, handling 1,400+ inquiries a day across Tamil, Telugu, Hindi and English, with an auto-translating WhatsApp RFQ agent mapped to Tally inventory.

Dreamwell delivers AI agents as a managed subscription. An agent is never “finished”: we monitor how it performs on real conversations and keep optimising it, so it keeps getting better as your products, prices and customers change. Pricing is shared after a free consultation.

How to ask for your own

You do not need to write a specification. Create AI Agents for my business asks a few ordinary questions — what you do, where customers write to you, what eats the day. Or WhatsApp +91 90805 83767 with a simple note: “Tiruppur knitwear, buyer enquiries and stock, AI agent venum.” That is enough for a first conversation in English or Tamil.

Sources

  • The Economic Times (2025), on Coimbatore textiles — Tamil Nadu textile export value, citing government data.
  • Mint, September 2024 — industry bodies on labour shortages and wages in Tamil Nadu textile and garment units.
  • James Oldroyd, Kristina McElheran and David Elkington, “The Short Life of Online Sales Leads”, Harvard Business Review, March 2011.

Solution-Based Services

How an AI agent helps textiles & knitwear →

Want this described for your own business?

Answer a few simple questions and we’ll recommend a first AI agent — or chat with us on WhatsApp. No jargon, no pressure.

Free consultation

Live in about 2 business days

Create AI agent →