Before building anything, I need a map. This entry lays out the entire AI finance landscape as I see it — which tools exist, where the gaps are, what's overhyped, and which systems I'll tackle first. This is the strategic roadmap.
Before I write a single line of code or test a single tool, I need to understand the terrain.
The AI Finance Landscape in 2026
The finance AI space is exploding. But most of what's out there falls into a few categories: - Bookkeeping automation (Dext, Hubdoc, receipt scanning) - Copilots that sit inside existing tools (Microsoft Copilot, Google Duet) - Standalone AI accounting platforms (trying to replace entire workflows) - LLMs being used as general-purpose assistants
What's missing? Systems that are purpose-built by finance professionals who actually understand the workflows, the edge cases, and the judgment calls that matter.
The Five Systems I'll Build
After mapping everything out, here's my plan of attack:
- AI Purchase System — Automating procurement from requisition to PO approval
- AI Invoicing System — Intelligent invoice capture, matching, and payment workflows
- AI Reporting System — Automated financial reports, variance analysis, and board packs
- AI Finance Strategy — Forecasting, scenario planning, and strategic recommendations
- AI Finance Function — Tying it all together into a cohesive, AI-powered finance department
How I'll Prioritise
I'm starting with the areas that have the highest manual effort and the most structured data — invoicing and reporting. These are the low-hanging fruit where AI can demonstrate clear, measurable impact.
Strategy and the full finance function come later — they require the foundation to be in place first.
What Success Looks Like
By the end of this journey, I want to have built working prototypes for each system, documented every lesson learned, and created a resource that any finance professional can follow to start their own transformation.
The direction is set. Now it's time to move.