🏠︎ | Past Sessions | Rethinking Finance Automation Enabling Strategic Finance Teams
- Event: Finance Forum 25
- Date: 7 October 2025
- Speaker: Mitja Sadar, CFO, Banxware
- Estimated read time: 7-8 minutes
Quick summary
Finance automation is not about removing people from finance, it is about removing low value work.
This session reframed automation as a design choice, not a technology decision. Mitja Sadar set out a practical way to separate what should be automated, what should stay manual, and what should sit in between with human oversight.
The core message was clear. Repetitive, rule based finance work should be automated aggressively. Bespoke, infrequent and judgement heavy work should not. The real prize is not efficiency alone, but freeing finance teams to focus on insight, decision support and strategic contribution.
Automation starts with redefining what finance actually does
The session opened by widening the definition of finance beyond accounting and reporting. Sadar framed finance as a combination of accounting, FP&A, treasury and corporate finance, each with very different automation potential.
Accounting captures and records activity. FP&A monitors performance and supports planning. Treasury manages cash, risk and liquidity. Corporate finance covers fundraising, M&A and investor interaction. Treating these areas as one uniform function leads to poor automation decisions.
Automation only works when it is matched to the nature of the work. Tasks that are frequent, predictable and rules driven are fundamentally different from those that require judgement, context and accountability.
A simple rule of thumb for finance automation
Sadar offered a clear test for automation decisions.
If a task is repetitive and does not require thinking, it should be automated.
If a task is infrequent, bespoke or judgement heavy, it should remain manual.
Everything else should be streamlined, with automation handling preparation and humans retaining sign off.
This framing avoids the common trap of trying to automate everything. It also avoids under investing, where teams accept manual effort as inevitable.
The goal is not full automation, but the right balance between machines and people.
What accounting should automate, and where to stop
In accounting, many core processes are mechanical. Bank reconciliation is a clear example. Matching transactions is not a thinking task, and machines can do it faster and more accurately than humans at scale.
Invoice processing can also be largely automated. Rule based systems can correctly classify the vast majority of invoices. The final few percent of exceptions are often not worth automating, as the cost of perfection outweighs the benefit.
Accruals are another strong candidate for automation. Once rules are defined, machines can accrue, release and post without manual intervention.
Accounts payable and receivable illustrate where human judgement still matters. While classification and processing can be automated, approval thresholds and fraud detection benefit from human oversight. The right threshold depends on company size and risk appetite.
Tax, precision matters more than speed
Tax shows why not everything should be automated. Routine filings such as VAT or sales tax are rule based and well suited to automation.
By contrast, tax optimisation, transfer pricing strategy and jurisdiction analysis require expert judgement. These areas involve interpretation, risk and personal liability. Preparation can be automated, but decisions should remain manual.
Sadar was explicit that accountability cannot be delegated to machines. AI can assist, but it cannot take responsibility.
Treasury, automation with guardrails
In treasury, execution can be automated once approvals are in place. Payments, liquidity dashboards and cash pooling can all run automatically within defined policies.
Risk management follows a similar pattern. Defining risk appetite is a human decision. Monitoring exposure and acting within that appetite is mechanical and can be automated.
Liquidity forecasting benefits from automation, but rarely reaches full autonomy. Forecasts can be generated automatically, but surprises and one off events still require human adjustment and interpretation.
FP&A and reporting, automate preparation, not decisions
Plan to actual analysis, consolidation and cost centre reporting are well suited to automation once structures are defined.
Budgeting is different. In most organisations, budgets reflect management judgement about markets, growth and risk. Automation can prepare inputs, but cannot make those calls.
Statutory reporting and audits follow the same pattern. Data preparation can be automated, but adjustments, interpretation and audit interaction remain manual.
Corporate finance remains fundamentally human
Corporate finance is the least automatable area. Investor communication, fundraising, M&A selection and post deal integration are bespoke by nature.
Templates and checklists help, but judgement, persuasion and coordination cannot be automated without losing value.
Automation supports corporate finance by reducing preparation time, not by replacing decision making.
Why finance automation really matters
The second half of the session focused on outcomes rather than processes.
Automation enables finance teams to move from doing work to shaping decisions. Real time visibility replaces delayed reporting. Scenario modelling becomes possible during meetings rather than weeks later.
“Automation enables us to go from doing the work to shaping where the business goes next.” Mitja Sadar, CFO, Banxware
With automation in place, finance can respond at the speed of the business. Teams can test scenarios, support strategy discussions and act as true partners to leadership.
Automation also improves consistency. Machines do not get tired, distracted or overwhelmed. At scale, this reduces error rates and improves trust in the numbers.
Most importantly, automation gives finance teams time. Time to think, challenge and add value.
What good looks like, practical actions for finance leaders
Questions to ask your team
- Which tasks are repetitive and rules driven, and still manual
- Where are humans spending time checking work that machines could prepare
- Which approvals genuinely reduce risk, and which only add delay
- Are we designing processes around accountability, or habit
- Finance is asked for insight during decisions, not after
- Reporting is near real time rather than retrospective
- Teams spend more time analysing than reconciling
- Automation rules are reviewed regularly, not set and forgotten
- Chasing full automation at the expense of judgement
- Automating exceptions instead of addressing the core workload
- Treating AI as a replacement for accountability
- Forgetting to redesign processes around the technology
What good looks like in practice
A strategic finance team automates the predictable, safeguards the risky, and focuses human effort where judgement matters. Automation becomes an enabler of better decisions, not an end in itself.
Conclusion, automation is a leadership choice
This session made a clear case that finance automation is not about tools. It is about clarity on what work matters.
When finance leaders make deliberate choices about what to automate, they unlock capacity, reduce risk and elevate the function’s role in the business.
The real transformation is not efficiency. It is relevance.
Speaker
CFO, Banxware, specialising in building investor ready finance functions and using structure, strategy and capital to support growth.