Fix the Semantic Model First: Lessons on Power BI, Snowflake and AI
I just watched Snowflake's Power BI webinar and it mirrors my client work: AI and BI only succeed when the semantic model is right. Daisy-chaining is a trap.
Exploring data analytics, AI trends, and visualization best practices
I just watched Snowflake's Power BI webinar and it mirrors my client work: AI and BI only succeed when the semantic model is right. Daisy-chaining is a trap.
Why LLMs and AI agents need governed semantic models and metric layers to eliminate hallucinations, enforce business logic, and deliver trustworthy enterprise insights.
Prompt engineering is fading. Context engineering is becoming the real operating layer for scalable, reliable, and governable AI systems.
Why healthcare is emerging as the proving ground for fine-tuned frontier models, and what this shift tells us about the future of personal, on-device AI.
Why building effective AI systems now depends less on clever prompts and more on how we design, curate, and manage context.
In 2026, data visualisation moves beyond dashboards into AI-assisted decision systems, where agents, semantics, and trust shape how insights are created, shared, and acted upon.
Reflections on my recent World Data Summit virtual roundtable about how data visualisation, storytelling, and AI come together to inform, inspire, and create meaningful impact.
How adopting global investor relations data and communication standards and applying effective data visualisation principles can build trust, transparency, and engagement across financial institutions.
Learn how to transform analysis into compelling data stories using a simple three-act framework — Onset, Development, and Outcome — to help your audience understand, believe, and act.
Augmenting Our Ability to Generate Visual Data Narratives. Artificial Intelligence is reshaping the way we approach data storytelling. This presentation explores how AI is diminishing the time needed for context-setting and analysis, making unstructured data more accessible, and elevating prompting as the new must-have skill.
Explore how inclusive design principles can enhance your dashboards and data visualisations for all users—combining accessibility with effective storytelling.
Discover the leading data visualization platforms of 2025, their standout capabilities, and how to select the right tool for your organisation.
Exploring how artificial intelligence is transforming the way businesses analyze and interpret data for strategic decision-making.
Learn essential principles for creating compelling data visualizations that drive business decisions and communicate insights effectively.
Discover emerging trends in data visualization including AI-powered insights, interactive storytelling, and immersive analytics experiences.
Discover emerging AI-driven trends transforming data visualization by 2025.
Create better PowerBI designs and solutions to help you communicate and connect with your audience.
Exploring the fine line between clarity and clutter in data visualization for more effective data storytelling.
Generative AI transforming data science, analytics, and visualization with LLM-powered agents.
Elevating marketing with data science: segmentation, recommenders, RFM analysis.
Enhancing data teams and leaders with design, communication and storytelling techniques.
Turn complex ideas and data into stories that resonate with your audience.
This is a bar chart race animation of daily COVID19 cases created using D3.js and Observable.
This post shows how easy it can be to geocode data using a custom script with google apps script in Google Sheets.
Mocking up data for analytics, datawarehouse or unit test can be challenging. This Python package is a fast and easy way to generate fake (mock) data.
An interview with Mike Sager, Vice-President of Customer Success of SiteMinder.
Highlights from the ACIC18 presented by The Sydney Informatics Hub at the University of Sydney
Slides for a talk I gave at Academy Xi on how to get into data science and analytics.
What are the key capabilities of a successful data science and analytics professional - now and in the future’? What are the things I should learn if I want to become a really good data science and analytics professional?
Three key learnings from a recent analytics project in partnership with The Works, a leading creative agency.
Joining me at Academy Xi are experts from leading organisations who will share their perspectives on all things big data and analytics.
The DASAMA™ model has been developed to help organisations gain visibility on their data science and analytics journey.
In this blog post I wanted to build a simple solution to help online shoppers decide on the best size of clothes they should buy while online.
While all of these elements are non-data related, I'd argue they can have as much of an impact on your data analytics initiative as issues relating to data quality, choice of analytical tool, time spent fine-tuning algorithms, etc.
This is a mobile-friendly interactive dashboard created in Tableau looking at some data from NAPLAN 2017 Preliminary results.
Data that are obtained in series of points over an equally spaced period of time are generally referred to as Time series data.
This is a mobile-friendly interactive dashboard looking at data for holders of visa subclass 457 over the past 6 years (2010-2016) in Australia.
CAGR measures the mean growth rates of money or units / quantities of something over the years. This is a simple function in Python to calculate CAGR.
Principal Component Analysis (PCA) is a popular method used in statistical learning approaches.
This is a bunch of slides I put together for a talk I gave recently on data visualisation and storytelling in the context of analytics.
Linear regression models are a key part of the family of supervised learning models.
K-Means is a clustering approach that belogs to the class of unsupervised statistical learning methods.
This is a visualisation created using Plotly.
When we study approaches for predicting qualitative responses, we study classification problems.
This visualisation is a product of the data wrangling work done in a previous post.
Today we are going to deal with some Microsoft Excel spreadsheet data in R.
Thoughts and a simple approach to a Kaggle competition using Random Forest in Parallel
Resampling is a technique that allows us to repeatedly draw samples from a set of observations and to refit a model on each sample in order to obtain additional information.
In this example we'll extend the concept of linear regression to include multiple predictors.
Linear regression is a type of supervised statistical learning approach that is useful for predicting a quantitative response Y.