In Gujarat's ceramic tile manufacturing cluster, most first contact with a customer happens on WhatsApp, not through a formal enquiry form. A dealer or a designer sends a photo — sometimes a screenshot from their own site — and asks for the same design: what size, what finish, and what it costs. Matching that photo to the manufacturer's own catalogue, then drafting a quote, is manual work that competes with everything else on a manager's phone that day.
Mirai360 AI builds an agent for exactly this workflow: photo in, catalogue match and drafted quote out, with a named person approving the quote before it goes back to the dealer.
How does the agent match a WhatsApp photo to a catalogue?
The agent reads the incoming image against the manufacturer's own catalogue — size, finish, and glaze — and proposes the closest match, along with the applicable price from the manufacturer's own pricing rules. The draft quote goes to a named person for review. Nothing is priced or sent without that review.
This is closer to visual search than to reading a document. A formal request for quote arrives as a spec sheet with numbers already on it; a WhatsApp enquiry arrives as a picture with no numbers at all. The agent's job is to supply the match a catalogue-literate employee would make by eye, at the speed WhatsApp already moves.
Why does a person approve every quote before it is sent?
A quote sent to a dealer is a commitment on price and specification. An agent that matched the wrong tile, or applied the wrong finish surcharge, would turn a fast reply into a wrong one — and a wrong quote costs more in a relationship-driven trade than a slow one does. Mirai360 builds the approval step into the workflow itself: we build your first business agent for one workflow, free — you approve everything it does. The agent drafts the match and the quote; a person with pricing authority reviews both against the actual catalogue before anything goes back to the dealer.
How is this different from a standard RFQ quoting agent?
Mirai360 also builds agents for spec-sheet-driven RFQ quoting (see our RFQ and quote generation post), where a buyer sends a formal spec sheet and pricing rules apply against known dimensions. The tile workflow starts earlier in the funnel: an inbound photo on WhatsApp, no formal spec, and a catalogue match that has to happen before any price can be quoted at all. Both agents end the same way — a person approving a drafted quote — but the tile workflow's first step is visual matching, not document parsing.
What else does a WhatsApp-native agent handle for a tile manufacturer?
Photo-to-quote is the first workflow, not the only one. The same approval-first design extends to routine dealer questions that already arrive on WhatsApp — a dealer asking when an order dispatches gets an answer drafted from the manufacturer's own dispatch data, with a person approving it before it sends.
How does Mirai360 apply its deployment discipline to a tile manufacturer?
The discipline is the same one Mirai360 applies to every workflow, because the industry-wide failure pattern is the same one: Gartner expects over 40% of agentic AI projects to be cancelled by the end of 2027, citing escalating costs, unclear value, and weak controls (Gartner press release, 25 June 2025: gartner.com).
- Pick one workflow first. Photo-to-quote on WhatsApp, not the whole sales process.
- Record a baseline before go-live. Current reply time and quote turnaround, written down before the agent starts.
- Keep a person on every quote. The agent drafts the match and the price; a named person approves before it sends.
- Log everything. Every match, draft, and approval sits in one audit trail.
- Scale against pre-agreed numbers. Wider catalogue coverage or wider agent authority follows only when the recorded numbers support it.
What does a tile manufacturer need to start?
A digitised catalogue with sizes, finishes, and current pricing rules; access to the WhatsApp number the enquiries already arrive on; and a decision about who approves a quote before it sends. No in-house data science team is required — the work on the manufacturer's side is describing the catalogue and the pricing rules it already has.
How do you measure whether it is working?
Against a baseline recorded before go-live: time from photo received to quote sent, the share of quotes that need a correction after drafting, and the share of enquiries answered same-day. Review on a set schedule, and widen the agent's role only when the numbers support it.
Frequently asked questions
What does the WhatsApp quoting agent actually do?
It reads an incoming photo, matches it against the manufacturer's own catalogue by size, finish, and glaze, and drafts a quote from current pricing rules. A person reviews and approves the quote before it goes back to the dealer.
Will the agent send a price to a dealer on its own?
No. Every quote waits for a named person with pricing authority to approve it before it sends.
Does this replace a standard RFQ quoting agent?
No, it complements it. RFQ quoting starts from a formal spec sheet; this workflow starts from an inbound photo with no spec attached, and both end with a person approving the drafted quote.
What does a tile manufacturer need before starting?
A digitised catalogue with sizes, finishes, and pricing rules, access to the WhatsApp number enquiries arrive on, and a decision about where approval sits.
Talk to us
Mirai360 AI works with tile manufacturers in Morbi and across Gujarat to scope a photo-to-quote agent against their own catalogue and baseline numbers, before anything goes live. We build your first business agent for one workflow, free — you approve everything it does. Book a 30-minute call: https://calendly.com/shivang-mirai360/30min
FAQ
- What does the WhatsApp quoting agent actually do?
- It reads an incoming photo, matches it against the manufacturer's own catalogue by size, finish, and glaze, and drafts a quote from current pricing rules. A person reviews and approves the quote before it goes back to the dealer.
- Will the agent send a price to a dealer on its own?
- No. Every quote waits for a named person with pricing authority to approve it before it sends.
- Does this replace a standard RFQ quoting agent?
- No, it complements it. RFQ quoting starts from a formal spec sheet; this workflow starts from an inbound photo with no spec attached, and both end with a person approving the drafted quote.
- What does a tile manufacturer need before starting?
- A digitised catalogue with sizes, finishes, and pricing rules, access to the WhatsApp number enquiries arrive on, and a decision about where approval sits.