AI agentic finance for finance teams is in progress and coming soon: AI agents that support FP&A, budgeting, forecasting and controlling workflows, with people approving every output.
Gert Jan van Straalen
Alongside my interim and consultancy finance work, I work with AI agents to improve financial processes within companies: the repetitive, rules-heavy work around reporting, reconciliations, budget submissions and variance commentary. I start from twelve years of knowing how those processes run in practice, in the insurance industry at Hiscox, and I keep a person in control of every number.
Most of the value in finance automation comes from understanding the process: where the data comes from, which checks matter, who signs off, and what the auditors will ask. That is the part I know well.
At Hiscox I redesigned and streamlined the processes for Budget & Reforecast submissions.
I managed the delivery of a new automatic reconciliation tool as part of the Finance Transformation Programme, and earlier transformed and accelerated processes for reconciling internal systems.
Monthly control of expenses against budget and reforecast, accruals, National Bank and tax declarations: the kind of work where accuracy and an audit trail come first.
Three ways to start, kept deliberately small. These describe what I offer; they are not results from earlier client work.
We look at one finance process together, such as month-end, the budget submission cycle or the monthly reporting pack, and map which steps an AI agent could prepare, which need human judgement, and where the controls sit.
A time-boxed pilot on one contained task, run alongside your existing process so nothing depends on it until it has proved itself. Typical candidates:
Combined with an interim or consultancy finance role: I bring AI agents into the day-to-day work where they help, and document how they are used so your team can run them without me.
Who owns it, what the inputs are, which checks matter and what the timetable is, before any tool is chosen.
An agent prepares, a person reviews and signs off. Every output can be traced back to its source data.
Agree up front what data the agent may see, where it is processed and which tools your IT and data-protection rules allow (GDPR).
Agree what a useful result looks like before starting, run the pilot in parallel, and then scale it, adjust it or stop.
I am open about this: the finance experience is long, the AI-agent experience is new and growing.
Twelve years in financial services: accounting, treasury, reconciliations, budgeting, forecasting and controlling, including process redesign and the delivery of an automatic reconciliation tool. Advanced Excel and VBA, Oracle ERP and Planning.
Working hands-on with AI agents to improve financial processes, applying them to the kind of finance work described above.
Details to confirm Which tools, courses or own projects to name here.
Running first small process reviews and pilots with companies that want to try this carefully. Case studies will be added here only once there are real results that a client has approved for publication.
Tell me about one process that takes too long. I'll give you an honest view of whether an AI agent can help, and what a small pilot could look like. For interim and consultancy finance work, see the finance services.
linkedin.com/in/gertjanvanstraalen
LocationBased in Lisbon, Portugal · working remotely with teams across Europe and worldwide · English, Dutch, German and Portuguese
