About The Formulator
PhD-level expertise, applied to your data
We combine deep scientific knowledge with modern machine learning to help manufacturers understand and improve their processes — without the overhead of a full-time data science team.
Why we exist
Most manufacturers collect far more process and quality data than they ever analyse. The data contains answers to questions about batch variation, yield loss, and product consistency — but extracting those answers requires a combination of statistical expertise, domain knowledge, and time that most teams do not have.
The Formulator was founded to close that gap. We work with manufacturers in chemicals, coatings, adhesives, cosmetics, and pharmaceuticals to turn their existing data into actionable process improvements — using machine learning methods that are rigorous, interpretable, and grounded in the science of the process.
Every analysis we deliver is reviewed by a PhD specialist before it reaches you. We do not produce black-box outputs. We produce specific, testable hypotheses and clear recommendations that your team can act on.
How we work
Data audit
We review your existing batch records, process logs, and quality data to understand what you have, what is missing, and what questions the data can answer.
Structured analysis
We apply machine learning methods — gradient boosting, random forests, SHAP analysis — to identify the process parameters and interactions that drive your quality outcomes.
Expert review
Every finding is reviewed by a PhD specialist with domain expertise in your industry. We translate model outputs into scientifically grounded, defensible conclusions.
Actionable recommendations
We deliver specific, prioritised recommendations — process parameter targets, raw material specification changes, monitoring improvements — that your team can implement immediately.
Our credentials
PhD specialists
All analyses reviewed by subject-matter experts with doctoral-level training in chemistry, chemical engineering, or statistics.
Industry experience
Direct experience in pharmaceutical, cosmetic, coatings, adhesives, and specialty chemical manufacturing.
Regulatory familiarity
Familiar with EU Cosmetics Regulation, ICH Q guidelines, ISO 17025, and GMP documentation requirements.
Interpretable methods
We use explainable ML methods (SHAP, partial dependence plots) so every finding can be understood, challenged, and defended.
What we believe
Rigour over speed
We would rather take an extra week to get the analysis right than deliver a fast answer that does not hold up under scrutiny.
Transparency over complexity
Every recommendation comes with an explanation. You should understand why we are recommending it, not just what we are recommending.
Your data, your insights
The findings we produce belong to you. We do not retain your data or use it to train models for other clients.
Ready to get started?
If you have batch records and a quality problem you have not been able to solve, we would like to hear about it.
Get in touch