AI & Automation
My special focus: putting AI and automation to work in finance — agents, workflows, forecasting and copilots — so teams spend less time on process and more on decisions.
The Big Picture
Data flows in, automation cleans and reconciles it, AI does the heavy lifting on forecasting and drafting — and a person reviews before anything goes out. An illustrative flow, not a single product.
ERP, Excel, PDFs, bank & market feeds
Pipelines that validate, match and reconcile
Models & LLMs that forecast, classify and draft
Judgment, sign-off and controls stay with people
Reports, dashboards and actions the business uses
Where It Helps
Not every task needs AI — often plain automation does the job. These are the areas I find most worth the effort.
The monthly grind — matched, checked and packaged on a schedule instead of by hand.
A first draft of the "why" behind the numbers, written from the data — edited by the person accountable for it.
Forecasts that refresh as new actuals land, with scenarios generated on demand.
Invoices, contracts and statements turned into clean, validated data.
Ask questions of your own data in plain English and get answers you can trace back to the source.
Automation that makes controls stronger, not weaker — continuous checks with a clear trail.
Under The Hood
AI drafts and flags; people decide and sign off. Automation removes effort, not accountability.
Every automated step leaves a trail — inputs, outputs and who approved — so it stands up to audit.
Sensitive financial data stays inside approved systems, with access limited to the people who need it.
Automate what saves real time or reduces real risk — and track it, rather than automating for its own sake.
Toolbox
A mix of data tools, automation platforms and AI building blocks.
Try It
Pick a recurring task your team does by hand and slide the numbers to match. It's simple arithmetic — a way to size an idea, not a promise.
Assumes 46 working weeks and a 40-hour week. Excludes build and run costs, so treat it as an upper bound.
Approach
Write down what actually happens today, step by step, before touching any tool.
Rules-based automation first — it's cheaper, faster and easier to audit than AI.
Use models for the fuzzy parts — reading documents, drafting, forecasting — with review built in.
Measure time saved and errors avoided, then iterate on what's working.
Color Theme
Paper (Default)