Closed-loop data
Not just defect labels — station actions and their outcomes: nest changes, mould recipes, lasting force, adhesive timing, rework and returns.
The world makes 24 billion pairs a year, mostly by hand
A sneaker is 40 to 60 components of leather, textile, foam, rubber and plastic that must be cut, stitched, moulded, lasted and bonded into a load-bearing, comfortable object — on lines that have barely changed in decades. Footeon is building the autonomy layer that changes them.
Footwear is craft-bound, waste-heavy and running out of hands
Leather yield, stitch tension, lasting fit, mould density and adhesive bond drift constantly with material lot, humidity, temperature and operator.
Irregular hides waste 15 to 30% of the most expensive material in the shoe. A weak bond delaminates in months and comes back as a warranty claim. Labour is scarce, rising and reshoring. Brands demand constant new models and colourways that explode changeovers. And the pattern cutters, lasting operators and sole technicians whose hands govern quality are an ageing, scarce craft. Meanwhile most plant managers are blind to real-time yield, bond quality, fit accuracy and defects across their own tables, lines and presses.
- 15–30% material waste on irregular hides
- Delamination is the dominant warranty failure mode
- High-mix launches multiply changeover cost
- The craft that holds quality together is retiring
The five make-or-break steps
- 01 · cut and nest — yield decided here
- 02 · stitch the upper — tension and seam quality
- 03 · mould the sole — density, geometry, cure
- 04 · last the upper — fit decided here
- 05 · bond the sole — warranty decided here
Each one gets a Footeon agent that perceives it and controls it.
- 09:14:02 hide.scan(batch="LH-2291", hides=64, sensor="line-b/cam-01")
- 09:14:19 64 hides mapped · 1,842 defects marked · mean usable area 91.2%
- 09:14:20 nest.optimize(pattern="velo-trainer-2", size_curve="EU36-46", objective="yield")
- 09:14:33 nest v7 · leather yield 87.4% (+8.3 pts vs baseline) · solve 12.3 s
- 09:18:44 cutter.execute(table="CT-2", nest="v7", blade_profile="leather-1.8mm")
- 09:22:56 2,412 parts cut · 0 out-of-tolerance · scrap 12.6% (plant baseline 20.9%)
Every line is real, copyable text and is announced politely to screen readers.
Autonomy belongs in the factory, not the dashboard
Footwear is materials-critical, fit-critical, waste-heavy, labour-heavy and high-mix. That is exactly the profile that rewards autonomy.
Our bet is that the winning product is not analytics on top of a plant but a system of action inside it — one that perceives every hide, upper and sole, decides the nest, the recipe, the profile and the window, writes those setpoints back through the machines, and takes responsibility for the result with an audit trail a brand auditor accepts.
- Trust before autonomy: grounding, audit, human-in-the-loop, graduated control
- Land narrow on one workflow, expand relentlessly
- Every supervised correction trains the system
- Vertical depth beats a horizontal dashboard
What we have built
Eight agents, a factory orchestrator, a factory-edge runtime and a shoe-and-line digital twin.
View as table
| # | Node | Agent | Depends on | Status |
|---|---|---|---|---|
| 01 | Twin simulate | Yield-and-Energy | — | SUCCEEDED |
| 02 | Cut and nest | Cut-and-Nest | Twin simulate | SUCCEEDED |
| 03 | Stitch upper | Stitch-and-Upper | Cut and nest | SUCCEEDED |
| 04 | Mould sole | Mold-and-Sole | Stitch upper | SUCCEEDED |
| 05 | Last upper | Last-and-Bond | Mould sole | SUCCEEDED |
| 06 | Bond sole | Last-and-Bond | Last upper | APPROVAL |
| 07 | Inspect pair | Quality-and-Fit | Bond sole | SUCCEEDED |
| 08 | Finish and pack | Robot-and-Assembly | Inspect pair | SUCCEEDED |
How we build
Four commitments that decide most of our arguments.
Trust before autonomy
We ship graduated control, not a black box.
- Citations on every output
- Human approval on validated parameters
- Immutable audit trail
- Fail closed, always
Honest numbers
A manufacturing claim that cannot be reproduced is worthless.
- Baseline before writes
- Measure per hide and per pair, not per batch
- Publish the method with the result
- Mark aspirational capability as aspirational
Vertical depth
Footwear, not generic manufacturing AI.
- Models per construction and material family
- Specs, lasts and adhesive datasheets as first-class knowledge
- Craft memory per cutter and technician
- Language the plant floor already uses
Edge first
The line cannot wait for the cloud.
- Inference on factory-edge hardware
- Standalone operation through outages
- On-prem where IP demands it
- Sub-second station decisions
Why now
Reshoring, labour scarcity, personalisation and sustainability mandates are all pushing the same direction.
- $18B total addressable market across footwear manufacturing software and automation
- $4.5B serviceable addressable market
- 13% annual growth in the segment
- 24B pairs of shoes made per year worldwide
Market figures are internal estimates built from public industry sources. [PLACEHOLDER] pending third-party validation.
Where we are going
Land the wedge, close the loop, then become the operations layer for the category.
-
0–6 months
Three to five design partners on the wedge workflow. MVP connectors, agent loop, review console and quality-grade audit log. Prove ROI in shadow then assist mode.
now
-
6–12 months
Convert design partners to paid. Graduated autonomy on lower-risk cutting and moulding. Twin hits fit, bond and yield before the run. SOC 2 Type II, SSO and RBAC.
next
-
1–3 years
Own the full cut → stitch → mould → last → bond loop end to end. Multi-site GTM through OEM and brand-sourcing channels. Benchmark data products.
scale
-
3–10 years
The autonomous operations layer for the world’s footwear manufacturing, with a platform and API ecosystem across every footwear category.
vision
What compounds
Every pair made on Footeon makes the next one better.
Network learning
Rare defect and delamination patterns learned at one plant transfer to plants running similar materials and constructions, IP-preserving. [ASPIRATIONAL]
Encoded craft
The scarce judgement of pattern cutters, lasting operators and sole technicians, captured per factory.
Qualification cost
Once a model is validated for a plant’s materials, lasts and adhesive system, replacing it means requalifying process control.
Channel
Cutting, moulding and lasting OEMs plus brand sourcing networks shipping Footeon-ready autonomy.
Twin fidelity
A simulation validated against millions of real pairs is not something a single-station tool can copy.
Facts and status
Stated plainly, including what is not yet true.
What Footeon is
- An independent startup building original Physical AI software and factory-edge autonomy
- Focused solely on footwear and sports-shoe manufacturing
- Pre-IPO and within the ten-year startup window
- Building models, runtime and twin in-house
What Footeon is not
- Not a consultancy, agency or systems integrator
- Not an internal division or a reseller
- Not a machine manufacturer
- Not a generic horizontal manufacturing dashboard
Who we build with
Design partners across three footwear categories.
“The nesting agent found yield our best pattern cutter could not — and then explained the layout hide by hide. We stopped arguing about scrap and started managing it.”
Leather yield 79.1% → 87.4%
“Bonding is where our warranty costs live. Having an agent watch humidity, primer flash-off and press pressure on every pair — and stop for a human when it wants to move a validated parameter — is the first thing that has actually moved delamination.”
Delamination returns −71%
“We run 40 model changeovers a month. Simulating the nest, the mould recipe and the bond window in the twin before the line starts took hours out of every launch.”
Changeover time −46%
Design-partner scenarios are illustrative and modelled on public footwear-manufacturing benchmarks — named references are [PLACEHOLDER] pending customer approval.
About the company
-
Athletic and sports footwear, casual and leather footwear, and safety footwear, across both brand-owned plants and contract manufacturers. Cut, stitch, mould, last and bond are shared mechanics; the models are tuned per construction — cemented, injected, vulcanised or direct-attach.
-
Only if you choose. Inference runs on factory-edge servers, and Enterprise deployments can run fully on-premises or in your VPC with no telemetry egress. Brand IP — patterns, lasts, construction specs — is tenant-isolated and never used to train models for another customer.
-
A typical wedge deployment on one line runs six to ten weeks: two weeks of sensing and baseline capture, three to five weeks of shadow-mode model tuning against your material lots, then graduated autonomy. Whole-factory rollouts follow line by line.
-
Line at $8,000 per line per month for one workflow, Factory at $55,000 per month for the whole loop, and Enterprise from $350k to $4M ACV for multi-site networks with custom material, last and bond models. Outcome-based components are available on leather waste, rework and returns.
Build the factory that makes every pair right
We are hiring engineers who want their code to move a machine, and partnering with plants that want to stop firefighting quality.