A collection of automation systems designed to eliminate repetitive tasks and build scalable business processes. From visual workflow builders to custom scripts.
Workflow automation in action: n8n pipelines, VBA macros, and process orchestration
The best automation is invisible. It's the report that arrives in your inbox every morning without anyone pressing a button. It's the data that syncs between systems without manual entry. It's the hours you get back to focus on work that actually requires human judgment.
Working in commercial finance, I see the same patterns everywhere: skilled people doing repetitive tasks. My approach is simple-identify the friction, understand the process deeply, then automate the mechanical parts so humans can focus on the strategic parts.
Visual workflow automation connecting APIs, databases, and services without vendor lock-in.
Custom scripts for data processing, file manipulation, and API integrations that run on schedule.
Excel/Office macros and scripts for repetitive tasks within business applications.
Power Query is where a lot of my data work begins. Rather than manually copying, cleaning, and reshaping data every reporting cycle, I build repeatable query pipelines that pull from multiple sources, apply consistent transformation steps, and refresh on demand. The result is reporting that stays accurate without anyone touching a single cell by hand.
The project I'm proudest of is a detailed weather model I built entirely in Power Query. It ingests raw weather feeds, normalises inconsistent timestamps and units, and joins historical readings against forecast data to surface the patterns that actually matter. Every step — from cleaning malformed records to deriving rolling averages and seasonality indicators — is codified in the query, so the whole model rebuilds itself the moment new data lands. What used to be a fragile, manual spreadsheet became a robust, self-updating analytical engine.
I've been embedding Microsoft Copilot across the business systems I work in — bringing AI directly into the tools my team already uses rather than bolting on yet another standalone app. The goal is to put insight and automation one prompt away from wherever the work actually happens.
Using Copilot to interrogate large datasets and reports in plain language, turning questions into answers without manual pivot-table archaeology.
Automating the first draft of recurring commentary, summaries, and emails so analysts review and refine rather than start from scratch.
Connecting Copilot to existing workflows so routine analysis and formatting happen inside the systems the business already relies on.
Reducing the lag between a question being asked and a confident, data-backed answer being available to the people who need it.
Every automation starts with watching someone do a task manually. Where do they sigh? Where do they copy-paste? That's where automation lives.
Automation that fails silently is worse than no automation. Every workflow includes error handling, logging, and alerting.
Automation should be maintainable by others. Clear documentation, sensible naming, and version control are non-negotiable.
Track time saved, errors prevented, and processes accelerated. Automation should have measurable ROI.
Measured Results
Automation isn't about being clever with code-it's about freeing up human capacity for higher-value work. Every system I build comes with before/after metrics: hours saved per week, error rates reduced, processes accelerated. The goal is always tangible business impact.