Consulting Today.
Orebody IQ Tomorrow.
We help mining companies understand not only what they know, but what they don't know. We are about turning data into knowledge and turning knowledge into decision confidence.
Mining does not merely need better ML models or better data science. It needs better ways of understanding the limits of its own knowledge.
Answers Create Decisions. Questions Create Understanding.
After decades working across geology, resource estimation, data management, and mining consulting, I repeatedly encountered the same problem. Projects focused on the data and the modelling, but the real challenge was often understanding what the data could and could not support and the various traps that cause projects to fail. It became that a new approach was needed. One that shifted from focusing on the algorithms to one that focused on making better decisions through asking better questions.
Mining is full of uncertainty. Geological uncertainty. Operational uncertainty. Economic uncertainty. Technical uncertainty. Most organisations respond by searching for answers. I start by asking questions. Because understanding rarely comes from having better technology. It comes from identifying the assumptions, gaps, risks, and opportunities hidden beneath the surface.
Asking different questions led me to looking at what is not known and looking at what caused projects to fail. Now I look for gaps in the data, for missing features, cognitive bias, semantic traps, and data traps. Missing requirements and missing specifications. Everything I do, I do to better define the problem space to make better decisions. From resource estimation and geometallurgy through to machine learning and Orebody IQ, I focus on one objective reducing uncertainty so better decisions can be made with confidence.
For years I thought I was building models. Machine learning models. Resource models. Geometallurgical models. Predictive models. Over time I realised the models were not the value. They were simply tools.
The real value comes from understanding. Understanding what is known. Understanding what is assumed. Understanding what is uncertain. Understanding where the greatest risk lies. The quality of a decision is often limited by the quality of the questions asked before it. That is why my work begins with questions, not algorithms.
Today my work combines geology, data science and software engineering to help mining companies turn data into knowledge and knowledge into confident decisions.
Why geology?
Geology is where I have spent most of my career. My expertise started with resource geology and it continues with the addition of data science and software engineering and product management.
Why mining data science?
While still at high school I had a strong interest in computer programming, 3D graphics, and AI. In early 2000s I got formalised training in computer programming and used those skills throughout my career as a resource geologist. Programming skills became highly valuable when I started using R in the early 2000s for statistical analysis as part of doing resource estimation projects. I have been using R for statistics and machine learning for decades. In essence I had been doing data science for decades. The combination of data science and geology has proven to be a potent combination for solving geology related mining problems.
What we deliver.
Understand the quality and fitness-for-purpose of your datasets.
Generate practical geological and geometallurgical insights from data.
Determine how much trust should be placed in a prediction, interpretation or recommendation.
What does a typical engagement look like?
A typical engagement starts with defining the project and the product (data solution) requirements. We first scope out the problem space and identify any potential traps and identify what is known and what is potentially missing in the data and determine if there is enough data to solve the problem. If there is a data solution then we move on to the next stage. From there a solution is designed and a specification writen. The requirements and the specification are then used to engineer the data solution (the implementation) in either R or Python or Go. The solution is then tested and validated against the specifications and then delivered as per the requirements.
Who is this for?
Geology Teams, Geometallurgy groups, Resource Geologists, Technical Services, Digital Transformation and Innovation Teams.
Why Nimmo Analytics instead of a traditional consulting engagement?
Nimmo Analytics has designed a workflow that maximizes the chance of success with the least amount of scope creep with the highest amount of value. Shifting risk left by ensuring the product (data solution) is fully engineered.
How is Nimmo Analytics different from traditional data science consulting?
Domain-first, requirement-first, question-driven, risk shifted left, emphasis on uncertainty, confidence before prediction.
Simplicity through structure
Focus on what matters. Remove what doesn't. Every analysis, workflow and decision should concentrate effort on reducing uncertainty, identifying knowledge gaps and improving decision confidence.
Clarity through understanding
Simplicity is not the absence of complexity. It is the result of understanding it. The objective is not to add more models, more analyses or more information. The objective is to identify what matters, remove what does not, and focus effort where it delivers the greatest improvement in understanding and decision confidence.
Evidence before models
Models should be supported by evidence, not assumptions.
Understanding before prediction
A prediction is only valuable when its limitations are understood.
Uncertainty is information
Uncertainty is not something to hide. It is something to quantify and communicate.
Knowledge compounds
Organisational knowledge should be preserved, accumulated and operationalised.
Decision confidence matters
The objective is not simply prediction. The objective is better decisions.
Decision first
The most valuable insight should never be hidden behind technical detail. Present the decision, confidence and key risks first. Provide supporting evidence and technical detail only as needed. The primary objective is enabling better decisions, not demonstrating analytical effort.
The frameworks are question-driven. Not solution-driven. Not technology-driven. Question-driven.
The 4D's Framework
What are the project goals?
What are the project requirements?
What assumptions are being made?
What is the desired outcome?
What form is the delivery?
What are the risks and contingencies?
What is the roadmap?
What is the proposed solution?
What are the solution specifications?
What is the acceptance criteria?
What data preparation is needed?
What testing/training framework will be used?
What are the steps?
What is the data quality?
What are the data gaps?
What data augmentation is required?
What ML algorithm is best?
What CLI commands are needed?
What is the final model?
What is the input/output format?
The Challenge Framework
What uncertainty exists?
What assumptions are being made?
What biases are influencing decisions?
What shortcuts are people taking?
What questions need to be asked?
What understanding is missing?
What can reduce uncertainty?
How do we know?
What supports the conclusion?
What validation exists?
What uncertainty remains?
What questions should be explored next?
What opportunities have emerged?
Why do you focus so heavily on questions?
Because most project failures are not caused by the model. They are caused by assumptions that were never challenged, requirements that were never defined, and uncertainties that were never identified. Questions expose these problems before they become expensive.
Reducing project risk.
Traditional data science is exploratory by nature. Mining consulting requires practical, defensible outcomes. The Nimmo Analytics framework was developed to reduce ambiguity, minimise cognitive bias and semantic traps, identify potential data traps, and structure analytical work within a repeatable engineering-oriented framework.
Continue the conversation.
My work extends beyond consulting engagements through public research, tooling and practical experimentation in mining data science, geology and orebody knowledge systems.
Industry commentary, project insights and professional updates.
Connect, follow, or DM.
LinkedIn ➔Research, ideas, playbooks, tools, datasets and practical resources exploring geology, geometallurgy, data science and orebody knowledge.
Start by visiting the HUB.
The HUB ➔Looking beyond analytics.
Throughout my career, I have been interested in a simple but important question. How confident are we in our understanding of the orebody?
Mining companies invest significant effort in collecting data, building models and generating predictions. Yet many important decisions ultimately depend on something less tangible: the quality of the knowledge supporting those decisions.
Orebody IQ is an evolving concept focused on understanding not only what is known, but also what is not known. It explores how organisations can identify knowledge gaps, assess knowledge confidence and better understand the limits of their current understanding. The long-term objective is not simply better analytics. It is improving decision confidence through a deeper understanding of orebody knowledge.
Explore Future Directions ➔Every project starts with a question.
Let's explore whether your challenge is a data problem, a knowledge problem, or something else entirely.
Book a Discovery Call