∇
Nabla Labs
High-tech physics‑based pattern grading

Most grading problems surface too late: in samples, in production, in returns.We catch them before you cut fabric.

Nabla Labs validates your existing graded patterns against a body distribution that matches your real customer base, and returns corrected DXF where grading breaks — upstream of sampling, before production.

Join early accessSee how it works
Simulation
● REPLAY
High-Precision, physics‑basedFit & Sizing

Problem

Sizing problems that are hard to see on paper.

The brands with the highest fit-related returns rarely have a factory problem or a grader problem. They have a grading visibility problem, one that compounds quietly across sample rounds and retail returns before it shows in the numbers.

Fit and sizing issues are estimated to drive 30–40% of fashion returns industry-wide. Most are set in motion before the first sample is sewn.
01

Inconsistent fit across the size run.

Fit decisions today rely on one fit model and one or two samples. Edge sizes, extended sizes, and shape diversity within a demographic stay untested until customers complain.

02

Returns and repeat sampling.

Grading errors set in motion months before production show up in sample rounds, retail returns, and repeated fit sessions — long after they can be fixed cheaply.

Grading Visibility Analysis
Gap detected

Single Fit Model

One body validates base size

Pass

Static Grade Rules

Same rule applied across all sizes

Size range output

XS
Untested
S
Assumed
M
Validated
L
Assumed
XXL
Untested
Blind spot: untested variance
Traditional Grading WorkflowSingle body · Single fabric · Static rules
1
2
2

Differentiation

What downstream tools can't see, and where the problem actually lives.

Size recommendation, virtual try-on, and fit advisor tools work downstream — they help a customer cope with fit issues the product already has. The fit problem they're managing was created months earlier, inside the grade rules. Nabla Labs works at that earlier stage. Same starting point as your existing fit work: your body standards, your size charts. Different unit of analysis: a population, not a point.

Body‑Standard Tooling Lives Here

Traditional fit standards define a representative body per demographic. Pattern is graded against that one body. Useful for setting brand‑level fit identity. Limited when customer shape diversity matters.

"What body should we design for?"
Our Engine Lives Here

We extend body standards with body distributions. Same starting point: your existing fit standards, your size charts. Different unit of analysis: a population, not a point.

"How does our pattern behave across all the bodies our standard represents?"
Read the population methodology
Closed‑Loop Correction
Most simulation tools stop at the problem. Nabla Labs is built differently: our closed‑loop simulation engine doesn't just surface where grading breaks, it computes what the corrected pattern geometry should look like. Your pattern maker receives a proposed corrected DXF alongside the diagnostic report. Not a black box. Not a replacement for their judgment. A precise, physics-derived starting point that would otherwise take hours of manual iteration to reach.

How it works

How it works.

You send a graded DXF/CAD pattern set and a description of your target customer. Nabla Labs runs a closed-loop simulation across a population of bodies that match that customer, identifies where grading breaks at each size, and returns a diagnostic report and an adjusted DXF.

01

Ingest

Existing DXF/CAD pattern files and grading specs. No re-modelling, no 3D-design seat.

02

Analyse

Physics-based simulation across a body distribution matched to your target customer, calibrated to your fit standards and fabric properties.

03

Diagnose

Per-size risk overview with strain and ease visualisation across critical functional zones — armholes, crotch, necklines, waistbands. The report flags which sizes are in spec and which need work.

04

Correct

Proposed corrected DXF geometry alongside the diagnostic. Not a black box, not a replacement for your pattern maker's judgment — a precise, physics-derived starting point that would otherwise take hours of manual iteration to reach.

Original Pattern Simulation
Issues Detected
MAX
MIN
Visualization of original fit issues
Before: Original Pattern
Excessive positive ease detected in critical areas
Optimized Pattern Simulation
Validated
MAX
MIN
Visualization of adjusted pattern
After: Adjusted Pattern
Balanced ease distribution and neutral balance

Outcomes

Results you can measure.

After a Nabla Labs analysis, your pattern maker has concrete notes on where to adjust, not just a feeling that something is off. Your fit sessions focus on the two or three sizes that actually need physical review, not a blind check of the full run. Your operations team has a documented fit standard that travels across factories and seasons. And you go into production knowing your edge sizes behave the way your base size does.

  • Fewer grading-related sample rounds per style.

  • Better use of fit model time, focused on sizes flagged as high risk.

  • A documented grading standard that travels across factories and seasons.

  • An optional Fit Quality Score per style: a documented, repeatable internal KPI you can track season over season.

SIM_LAYER: STRAIN_DISTRIBUTION
MAX
MIN
Strain distribution analysisCritical area visualization

Works with your existing pattern maker, grader, factory, and CAD workflow. No 3D seat, no GPU workstation, no change to how your team designs or grades.

Read the methodology

FAQ

Common questions

Requirements

Who is Nabla Labs for and what are the technical requirements?

Nabla Labs is built for technical designers, pattern makers, fashion brands, and grading houses. No 3D knowledge and no specialised software licences are required. The engine is a low-investment way to bring high-fidelity physics to your existing DXF/CAD workflow.
Integration

What is required to integrate Nabla Labs into my current process?

The bar for entry is intentionally low. Nabla Labs only requires your DXF/CAD pattern files and grading specs. The engine hooks into your existing workflow without requiring you to change how your team designs or grades patterns.
Differentiation

How does this compare to 3D tools like CLO3D or Style3D?

CLO3D and Style3D are excellent simulators, but you still need to interpret and act on those simulations manually. Nabla Labs does the interpretation and correction for you. If you already use 3D tools, Nabla Labs complements them; if you use DXF-only workflows, Nabla Labs gives you 3D-level insights without the overhead of a full 3D stack. If your team doesn't have the bandwidth to run and interpret 3D simulations in-house, Nabla Labs delivers the same level of insight as a report, without the tooling overhead.
Positioning

How is this different from Virtual Try-On (VTO) widgets?

VTO tools manage a symptom. The symptom is a customer who can't find their size. The underlying cause is a grade rule that makes XL fit differently from M, and that cause was set in motion during pattern development, long before the garment reached a product page. Nabla Labs works at that earlier stage: correcting the grading so the inconsistency that would have required a VTO workaround never gets produced in the first place.
Technology

Does this rely on Generative AI for the simulations?

No. While Nabla Labs uses AI for some optimisations, the core engine is based on high-fidelity physics. The engine simulates the mechanical interaction between fabric and body across diverse shapes and dynamic postures, and returns corrected grading and clear diagnostic reports.
Categories

Which garment categories do you support?

Nabla Labs supports virtually all garment categories. The engine focuses on critical functional zones (armholes, crotch, necklines) to identify where grading errors most frequently affect comfort and aesthetics, helping you iterate faster.
Body Modelling

How do you generate the bodies you simulate on?

Nabla Labs uses an internal body-shape model, calibrated against public anthropometric datasets and any size charts or fit standards you provide. The model can sample bodies that match a specific demographic (region, age band, athletic or standard build) or a custom distribution defined for your customer base. The output is a population of bodies suited to your brand's market, used as the test bed for your patterns.
Materials

How do you handle fabric physical properties?

Nabla Labs works with measured properties when available, or estimated properties from descriptions and reference samples. Fabric behaviour is part of the grading decision, not an afterthought. See the Fabric page for the full fabric-aware grading approach.

If you manage grading or fit, this is built for that problem.

Early‑access participants get access to the tooling, feedback on their own styles, and influence on how the diagnostics evolve.

Join early access →Talk to us →

Pricing per style. Volume and early‑access rates on request.