KevinTumen

AI & Finance

Towards an AI-augmented finance department.

I don't talk about AI as a spectator: I use it every day to build management tools that run in production. Here is what I've learned, and where I want to take it.

My practice

My AI education doesn't come from conferences but from the keyboard. Since 2025 I have developed, in tandem with AI, an ERP, a collaborative EPM and two websites, all in production on a server I administer. Before that, I had already automated the reporting chain of a 14-entity group: API flows between the ERP, accounting, CRM and HRIS systems, delivered through Power BI.

That practice taught me something the slide decks don't say: AI doesn't replace business expertise, it multiplies it. It executes fast, but you have to know precisely what to ask, and be able to judge what it returns.

My convictions

  1. The finance function can now build its own tools.

    For twenty years, FP&A has lived in the gap between what the ERP does and what the job needs, a gap filled with Excel. AI changes the equation: a finance professional who knows their processes can now specify and have built the exact tool they're missing, in weeks, not years.

  2. Closing speed is an architecture choice, not a team effort.

    I've lived the journey from WD10 to WD3: it isn't achieved by asking teams to stay later, but by eliminating manual collection. APIs, automated flows, embedded controls: AI radically accelerates putting all of that in place.

  3. Business partnering starts where production ends.

    Every hour saved producing the numbers is an hour available to explain them, challenge the assumptions and inform decisions. AI doesn't make the FP&A analyst obsolete: it finally frees them for the part of the job that carries value.

  4. Understanding the technical chain has become a competitive advantage.

    I don't believe in the finance leader who outsources without understanding. Knowing what an API, a database, a deployment actually are, not to do everything yourself, but to talk as an equal with IT and software vendors, cost a project realistically, and spot dead ends before committing budget to them.

The destination

An AI-augmented finance department.

My goal is clear and stated: take responsibility for a finance department and apply there what I have been building for more than ten years. Fast closings, reporting without manual collection, tools that serve the business, a team that spends its time analyzing rather than producing, and an FP&A function acting as a true business partner, informing the executive committee rather than compiling for it. A finance department that uses AI not as a slogan but as a daily operational lever, measured by the same standards as everything else: reliability, speed, cost.

Until then, I keep sharpening both blades: the craft, by supporting demanding multi-site environments; the tooling, by building and operating real management applications.

Do these topics speak to you?

AI applied to finance, performance management, career paths: always glad to exchange.

kevin@tumen.fr