QuoteWeaver
Measure, price and sign a curtain and blinds order in a single home visit.
The short version.
QuoteWeaver runs a soft-furnishings business end to end: leads, site measurements, quotes, orders, supplier purchase orders, work orders for the workshop, and delivery. Its centrepiece is a pricing studio where the shop owner sets up their own pricing logic once, without code, so an estimator can price any window on site in seconds.
The first customer was a window-furnishings company in Kochi that ran quoting from a large Excel workbook. The product grew from two small tools we built for them, a measurement form and a quoting sheet, into a multi-tenant SaaS that any furnishing business can sign up for.
What was broken.
In soft furnishings, one quote depends on fabric width, fullness, pattern repeat, heading and hem allowances, supplier size charts and square-foot slabs. Our first customer ran it all in a spreadsheet, and when we audited it we found totals that only calculated on the first row, a fixing charge left out of every blind quote, and a broken reference in the summary. Estimators measured on paper, priced back at the office, and quoted days later.
- Quotes took days, so customers compared prices and cooled off.
- Spreadsheet errors quietly under-charged jobs.
- Supplier price lists arrived as PDFs and phone photos and were typed in by hand.
From idea to production.
- May 2026
First tools, first tenant
Measurement form and quoting sheet replaced the spreadsheet, then became a multi-tenant web app with clients, projects and branded printable quotes.
- Jul 2026
Pricing engine designed
A three-layer pricing model (catalogue, recipes, quote rules) prototyped and tested against real quotes to the rupee.
- Jul 2026
Full platform build
Orders, purchase orders, work orders, delivery, reports, team roles and billing built in six weeks. A full acceptance run passed with hand-checked maths.
- Aug 2026
AI price-list import
Supplier PDFs and scans read straight into the catalogue.
What we built.
The owner describes their pricing once in the pricing studio: fabrics with their width, rate and repeat, product styles with fullness and making charges, supplier size charts, and quote-level rules for discounts and bundles. In the field, the estimator fills one form per window (room, furnishing, measurements, options) and sees the live price of every line. The customer accepts on a link before the estimator leaves, and the order, purchase orders and work orders follow automatically.
- Step 1Estimator measures each window
- Step 2Pricing recipe works out the fabric
- Step 3Live price per line with tax
- Step 4Customer accepts on a link
- Step 5Order, PO and work order created
What we delivered.
Pricing studio
Owners build pricing from blocks and formulas, referencing measurements, fabric rate, width and repeat, with size-chart lookups and if/then rules. Every formula shows a live preview as you type.
One-form site visit
Room, furnishing, measurement, options and live price, all on one screen. No wizard to click through while the customer watches.
Window measurement diagram
A drawn window with tap-to-edit dimensions for header, footer, sides, depth, height and width, matching the paper template estimators already use.
AI price-list import
Upload a supplier's PDF or a photo of a printed list. AI reads the table, maps the columns, and the catalogue fills in, without duplicates on re-import.
Customer quote portal
Customers review and accept quotes online, with order terms and a payment schedule.
The whole order chain
Quote to order to supplier purchase order to workshop work order to delivery order.
Reports
Profit and loss, pipeline and order reports for each business.
Built for more than one country
INR with GST, AED with VAT, and billing by region.
A closer look.
Intelligent, and trustworthy.
Read, then map
AI first copies the supplier's table exactly as printed, then a second small step works out which column is which from a few sample rows. Everything after that is ordinary, predictable code.
Maths is never left to AI
Prices, fabric consumption and taxes are calculated by a tested formula engine, never by a language model.
Any model, same result
The import works with a range of AI models and can switch between them without code changes. It was tested across models with a 100× difference in price.
Errors in plain language
If a formula can't be calculated yet, the studio says what is missing instead of showing an error code.
The stack.
- Platform
- Next.js 16 on Cloudflare Workers
- Data
- Neon Postgres with per-organisation isolation
- AI
- Gemini via OpenRouter for document reading
- Business
- Dodo Payments billing, PDF quotes, Resend email, Google sign-in
The numbers.
Build proof
Expected impact for users
Verified from project repositories, test runs and release records. Projected modelled estimates of user impact, not measured client results.
Got an idea? Let's ship it.
Tell us the workflow that's costing you customers or hours. We'll come back with a prototype plan, a timeline in weeks, and a fixed quote.