AI-driven defect tracing for global steel manufacturers

Minimise scrap rates. Maximise production quality. Improve your bottom line.

Join advanced steel manufacturers resolving defects with intelligent solutions tailored to their mills

Deep.Meta identifies problems in metals production by harnessing manufacturing data to predict when defects will occur, before they do. We give operators and managers clear insights that enable them to consistently deliver high-quality products with minimal scrap.

How it Works

We isolate critical data, and then train & deploy machine learning models tailored to your production line.

1| Outline

We outline your most pressing product problems and agree performance metrics.

2| Cleaning




During cleaning we resolve anomalous signals from sensors, so the data is accurate.

3| Analysis

Conduct exploratory analysis to visualise relevant trends in your production data.

4| Training

Models are trained with the data and become able to trace and predict defects.

5| Deployment

Models are deployed directly to your production lines as an easy to use software.

Unique Technology

Conceived from cutting-edge PhD research, Deep.Meta is the only truly data-driven solution for minimising your scrap rate.

Unparalleled precisions & rapid detection of defect sources

Automatic defect isolation saves 100s of hours per year

Tight production controls achieve target properties without expensive alloying elements

Our ML models learn continuously from their environment as your production line evolves

Book a in-demo today to find out more and talk about joining our beta programme.


Learn more about how Deep.Meta can help you optimise your production line.


In-depth, fundamental understanding of physical processes in metal production.


We use secure, private cloud services to deliver insights straight to the production line.


We trace the causes of specific metallurgical defects most important to our customers in real time.


Simple interface allows you to make production decisions using deep learning insights without being machine learning experts.

The Team

Two uniquely qualified founders.

Osas Omoigiade


Imperial College London

PhD Metallurgist

Developed novel process for global steel manufacturers

Aizar Enciso Dominguez


Imperial College London

PhD Quantum Computing

Led teams to build neural networks