Data Analysis & Insights.

I investigate business questions, identify patterns and relationships, and translate statistical results into findings that can support clearer decisions.

Discuss a project

Python R SQL Regression Hypothesis Testing PCA & Clustering Data Visualisation

THE PROBLEM

Having data is not the same as knowing what it means.

Tables and dashboards can describe what happened without explaining which patterns matter, what is associated with what, or how much uncertainty surrounds the result. I use exploratory and statistical analysis to turn raw observations into interpretable evidence around a concrete question.

WHAT I CAN ANALYSE

From descriptive patterns to relationships, segments and statistical evidence.

PERFORMANCE & TRENDS

Understand how an indicator behaves

Compare performance across time, categories, markets or locations and identify the changes, distributions and outliers that deserve attention.

RELATIONSHIPS & DRIVERS

Investigate what moves with what

Examine associations between outcomes and explanatory variables using regression, correlation, statistical tests and model diagnostics.

SEGMENTS & PROFILES

Find structure inside complex data

Group observations into interpretable segments and reduce high-dimensional information into clearer underlying patterns.

UNCERTAINTY & COMPARISONS

Put observed differences into context

Assess whether differences are meaningful, quantify uncertainty and compare alternative models, groups or analytical assumptions.

ANALYTICAL METHODS

The method depends on the question the data needs to answer.

REGRESSION & INFERENCE

Quantify relationships and test whether observed effects hold up.

Build and diagnose explanatory models, test statistical hypotheses and evaluate how selected variables relate to an outcome while keeping interpretation central.

OLS Regression Multiple Regression Polynomial Regression Hypothesis Tests Model Diagnostics

MULTIVARIATE STRUCTURE

Reduce complex variables into interpretable dimensions.

Explore latent structure, correlated variables and categorical relationships when a dataset contains more dimensions than can be understood one column at a time.

PCA Exploratory Factor Analysis Correspondence Analysis MCA Correlation Structure

SEGMENTATION & CLASSIFICATION

Discover groups or predict membership from observed patterns.

Use clustering when the groups are unknown, or supervised classification when labelled outcomes are available and the objective is to predict or explain class membership.

K-Means Hierarchical Clustering Model-Based Clustering Logistic Regression LDA Random Forest ROC / AUC

SAMPLING, TIME & DISTRIBUTIONS

Account for how observations were generated and how they evolve.

Incorporate sampling design, weighting, distributional behaviour or time structure when the analytical question cannot be treated as a simple cross-sectional comparison.

Sampling Weights Imputation Distribution Models Time-Series Analysis Trend Diagnostics

HOW I WORK

From a business question to evidence that can be explained and used.

DEFINE THE QUESTION

Start with the decision the analysis needs to support.

Clarify the outcome, comparison or uncertainty that matters and determine what the available data can reasonably answer.

EXPLORE THE DATA

Understand structure before fitting a model.

Examine distributions, missingness, outliers, relationships and group differences to determine which analytical approach is appropriate.

MODEL & VALIDATE

Apply the method and test whether the result is robust.

Fit the selected statistical or machine-learning model, inspect diagnostics and compare performance or assumptions where needed.

INTERPRET & COMMUNICATE

Translate output into a conclusion people can use.

Present the relevant findings, limitations and visual evidence in a concise form suited to the people making the decision.

TYPICAL PROJECTS

When the question requires more than a descriptive table.

MARKET & PERFORMANCE ANALYSIS

Turn operational or market data into a structured analytical picture.

Compare prices, activity, volumes or other indicators across periods, categories and locations, highlighting the changes and differences that matter.

EDA Descriptive Statistics Trend Analysis Visualisation

DRIVER & RELATIONSHIP ANALYSIS

Investigate which variables are associated with an outcome.

Build interpretable regression-based analyses, assess statistical evidence and inspect model diagnostics to distinguish stronger relationships from superficial patterns.

Regression Hypothesis Testing Diagnostics R

CUSTOMER OR PRODUCT SEGMENTATION

Find groups that are not explicitly labelled in the source data.

Combine dimensionality reduction and clustering methods to identify useful profiles, explain how groups differ and support more targeted analysis.

PCA K-Means Hierarchical Clustering Profiles

SURVEY & MULTIVARIATE ANALYSIS

Extract interpretable structure from many related variables.

Analyse survey or research datasets using weighting, factor-based methods, correspondence analysis or other multivariate techniques suited to the measurement structure.

EFA CA / MCA Sampling Weights Statistical Reporting

HAVE AN ANALYTICAL QUESTION?

Let’s turn it into something the data can answer.

Tell me what you are trying to understand, compare or explain. I can help translate the question into a concrete analytical workflow and a decision-ready result.

Discuss a project