The company

UK-incorporated. Focused on R&D software.

AUROLINK Innovations Ltd builds governed ML tools for materials and industrial R&D. Our first product screens alloy candidates and returns ranked, reviewable recommendations.

Who we are

We build tools for expensive experiments.

In materials R&D, the hard question is often not whether there are more ideas. It is which few candidates deserve the next round of testing. AUROLINK focuses on ranked recommendations, uncertainty, refusal on unsupported inputs, and clear provenance.

Our first product is a live high-entropy-alloy hardness screener. The same engineering pattern can extend to other properties and small-data R&D workflows where reviewability matters.

How we work

We keep the claims narrow.

AUROLINK is built for screening decisions, not qualification decisions. We show uncertainty, keep provenance, and record when a model is not good enough to promote. That restraint is what makes the output usable in review.

Where we're going

Start narrow, then expand with evidence.

We are starting with alloy screening because the use case is specific, valuable and testable. Expansion only makes sense where the data supports a useful tool.

Now

The alloy-screening product is available today. We are open for scoped custom builds and paid design-partner screening pilots.

Next

Retained screening and private models trained on customer process data where the data is strong enough.

Then

More alloy properties, adjacent material families, and a UK delivery team as engagements grow.

Company

The details.

AUROLINK Innovations Ltd is based at the Canterbury Innovation Centre, University of Kent. We build ML software for materials and industrial R&D and operate our own alloy-screening product.

Company
AUROLINK Innovations Ltd
Registered
England & Wales · No. 16883969
Incorporated
1 December 2025
Office
Unit 72, Canterbury Innovation Centre, University Road, Canterbury, Kent CT2 7FG
Focus
Governed applied AI · ML systems & products

Get in touch

Have a screening or ML problem?

Tell us what you are trying to decide. We will say what the data can support and what a sensible first engagement looks like.