Private beta By invitation

AI for rubber
chemistry.

We predict cure, hardness & more — straight from the recipe. A neural ensemble draws the curve below.
MDR @ 170°Cspec · NBR/EPDM 70:30predicted by neural ensemble · α 0.62
● predicting · 04:48 / 12:00 min
Torque · dN·m
22
16
10
4
measured · MDR
predicted · model
±σ envelope
024681012 min
predicted t90 = 4.8 min
scorch ts2 safely > 1.4 min

A platform that learns your elastomer chemistry and predicts how a compound will behave — before you mix it. Forward prediction from recipe and process, and inverse design from a target spec.

↳ we're onboarding a small group of design partners.
Request early access → See how it works
No commitment · we'll be in touch
/ what the platform does

Three things, done well.

Request a demo →
/01 · forward

Predict properties

Recipe and process go in; cure curve, hardness, tensile and more come out — with a confidence band.

/02 · inverse

Design to a target

Give it a target spec — hardness, compression set, cost — and it proposes candidate recipes to test.

/03 · private

Your data stays yours

Models learn across labs without ever pooling raw recipes. Isolated by design, opt‑in, revocable.

Built for real elastomer chemistry — not generic materials data
NR natural SBR styrene‑butadiene BR polybutadiene EPDM NBR nitrile HNBR CR chloroprene IIR butyl FKM fluoro silicone tri‑ & tetra‑blends + more families
Request form ~90 sec

Request early access.

We're working with a handful of R&D teams while the platform is in private beta. Tell us a little about your lab and we'll set up a demo.

  • A walkthrough demo tailored to your elastomers.
  • A look at forward prediction and inverse design on real compounds.
  • An NDA on request — your recipes never leave your hands.
Early‑access request
FORM · RA‑001

We'll get back to you personally. No sales pressure.

Your data is used only to contact you about the beta.
/ a few quick answers

Before you ask.

001Is the platform available now?+
Not yet — we're in private beta and onboarding a small group of design‑partner labs. Requesting access here puts you on the early list for a demo.
002Will our recipes and data stay private?+
Yes. Your raw recipes never leave your tenant and are never pooled into a shared model. Any cross‑lab learning happens through federated learning, opt‑in and revocable. NDA available on request.
003Which elastomers does it cover?+
It's built for real rubber chemistry across families — NR, SBR, BR, EPDM, NBR, HNBR, CR, IIR, FKM, silicone and more — including blends. We'll tailor the demo to the compounds you actually run.