About the role
What Exactly Should You Look For?
Core Expertise: Python, Machine Learning, Time-Series Modelling, and Numerical Optimisation.
Preferred Industry Background: Manufacturing, Metals, Mining, Cement, Chemicals, Oil & Gas, and other Process Industries.
A. Strong Python & Machine Learning Expertise
Must-have skills:
- Strong proficiency in Python programming.
- Experience in Machine Learning model development.
- Hands-on experience with LightGBM.
- Knowledge of GAM (Generalized Additive Models).
- Feature engineering.
- Model validation and diagnostics.
- Model explainability.
What to look for:
The candidate should be capable of reviewing existing ML models, identifying areas for improvement, and optimising their performance rather than building everything from scratch.
B. Time-Series & Industrial Process Modelling
This is one of the most important aspects of the requirement.
Must-have expertise:
- Time-series forecasting and modelling.
- Industrial process data analysis.
- Equipment performance modelling.
- Experience working with historian data and laboratory data.
- Understanding of lagged features and sensor data.
- Knowledge of process variables and operational parameters.
- Predictive modelling for manufacturing operations.
What does this mean practically?
The candidate should understand how different industrial variables interact and develop models that can predict final outcomes based on historical and real-time operational data.
C. Constrained Numerical Optimisation
Relevant skills:
- Numerical optimisation.
- Constrained optimisation.
- Metaheuristic algorithms.
- Grid Search.
- Optimisation of operating parameters.
- Multi-objective optimisation.
- Throughput, recovery, and cost optimisation.
- Sensitivity and scenario analysis.
What to look for:
Candidates should have experience developing optimisation approaches that identify the best operating parameters while considering industrial constraints, production targets, recovery rates, and operational costs.
D. Industrial Domain Experience
Preferred industry backgrounds:
- Manufacturing.
- Mining and Metals.
- Cement.
- Chemicals and Petrochemicals.
- Oil & Gas.
- Power and Energy.
- Industrial Automation.
Relevant experience:
- Working with process engineers, plant operators, operational SMEs, and industrial data.
- Understanding industrial processes and operational parameters.
- Experience in manufacturing optimisation or process optimisation.
Candidates with direct experience in industrial or process optimisation would be particularly relevant.
E. Model Pipeline & Live Testing
Look for candidates who have hands-on experience with:
- Machine Learning model pipelines.
- Model deployment and inference workflows.
- Kedro (QuantumBlack) desirable, but not mandatory.
- API integration and data exchange.
- Live model validation.
- Shadow testing.
- Model monitoring and feedback loops.
What to look for:
The candidate should be comfortable running predictions on live plant data, comparing predicted outcomes with actual results, validating model performance, and improving models through continuous monitoring and feedback.
Key Hiring Priority
The ideal candidate should combine three major capabilities:
- Machine Learning: Strong Python, LightGBM, GAM, feature engineering, and model diagnostics.
- Industrial Modelling & Optimisation: Time-series modelling, process data understanding, and constrained numerical optimisation.
- Practical Implementation: Model pipelines, deployment, live plant data validation, and shadow testing.
Important: Prioritise candidates who have applied Machine Learning and optimisation to real industrial processes rather than candidates with only theoretical ML or general data science experience.