Fabrication · anonymised
From drawing to quote: how a laser-cutting and bending shop automated its RFQ workflow
A flat and tube laser-cutting and press-brake fabricator in Western Canada with 15 to 40 staff and two estimators. The client’s name is withheld under NDA.
- Customers served
- Agricultural equipment makers, machinery and electrical businesses, furniture and shop-fitting, product developers needing prototypes
- Files received
- DXF, DWG, STEP, SLDPRT, IPT, PDF drawings and ZIP bundles, by email and web enquiry
- Engagement
- Discovery and file audit, then CAD processing and nesting services, then quoting workflow integration, delivered by a dedicated eSquall team working remotely with weekly reviews
- Delivered
- DXF and DWG reader, 3D import and sheet-metal unfolding, nesting service, and outputs as DXF, STEP, PNG preview and structured part data
The situation
Every customer enquiry arrived as an email with a mix of file types and a few lines of text. Before a price could be written, someone had to open each file in the right CAD tool, check that it was complete, measure the cut length and count the holes, unfold any bent parts, work out how many sheets the job would take, and type all of it into a spreadsheet that held the shop’s rates. The two estimators spent most of their day finding numbers rather than deciding prices.
The owner’s concern was not only cost. Slow quotes were losing work to shops that replied the same day, and two estimators pricing the same part could land 10 to 15 percent apart, because the rules lived in people’s heads rather than in a system.
What eSquall did
1 · Discovery and file audit
2 weeksWalked through 30 recent enquiries end to end; catalogued every file type and failure mode (3D DWG, missing units, open contours, native SolidWorks files); captured rate tables, bend tables and sheet sizes; agreed what a correct result means for each file
2 · CAD processing engine
8 weeksBuilt services that read DXF and DWG and extract part size, cut length, holes, bends, thickness and weight; import STEP, SLDPRT and IPT and produce flat patterns for suitable sheet-metal models; export DXF, STEP, PNG preview and structured JSON
3 · Nesting service
3 weeksArranged flat parts on the shop’s sheet sizes with spacing and kerf, returning sheets needed and utilisation
4 · Quoting integration
3 weeksFed extracted measurements and sheet counts into the quoting workflow so the estimator reviews and approves instead of re-typing; preserved the shop’s own pricing rules
5 · Validation
2 weeksCompared engine results with the estimators’ manual measurements on a reference set of real files; tuned tolerances; documented which file types are fully automatic and which need a review
What changed
One rule set
Every estimator prices the same way, and the rules are versioned instead of remembered.
Review, not re-typing
Measurements and sheet counts arrive in the quoting workflow; the estimator checks and approves.
Checked against the shop’s CAM
234 of 249 checks on the shop’s past parts matched its own CAM software: size, cut length, weight, pierces and parts per sheet.
What we learned, and built into the product
A supported extension is not a supported file.
DWG can hold 2D or 3D, and a STEP file may not be a sheet-metal body. The engine now classifies by content and tells the estimator exactly why a file needs a different export, with a one-click request email.
Confidence beats silence.
Every measurement carries a confidence score and warnings. Low-confidence parts go to review; high-confidence parts price automatically.
Pricing rules are the real asset.
Capturing feed rates, pierce times, bend tables and margin rules in a versioned rule set made the business less dependent on one person.
Cloud CAD dependencies are a cost and a risk.
Processing native CAD through a vendor cloud works but adds per-file cost and an external dependency, so the next version runs on our own geometry engine, with the vendor route only as an optional add-on.