From Excel to AI. How to automate cash flow in an SME
Building on the approach I describe in my book Finanse dla niefinansistów w praktyce (Finance for Non-Financial Managers in Practice), which places great emphasis on financial processes, financial modelling, tools for budgeting and cash flow, and the integration of systems, it is possible to build a semi-automatic or almost fully automatic process for forecasting liquidity.
The SME’s problem: the data are everywhere and hard to integrate
A typical company has:
an ERP system (Comarch ERP XL, Optima, Enova, SAP Business One, Microsoft Dynamics and so on), although a large proportion of SMEs rely on an outsourced accountant, who has one of the accounting systems,
Saldeo or a similar system for handling the flow of financial documents,
online banking,
a CRM,
Excel,
very rarely, Power BI.
As a result, every day the owner asks:
“How much money do we have?” and not: “How much money will we have in three weeks’ time?”
Automation model, Level 1 – Excel + Saldeo
The simplest stage, in which we automatically pull in:
From Saldeo:
sales invoices,
customers’ payment dates,
cost invoices,
suppliers’ payment dates.
From the bank:
account balances.
Excel or Microsoft 365:
calculates a forecast three weeks ahead.
In this model, the accounts department no longer prepares the report by hand. Updating it takes a few minutes.
Model, Level 2 – ERP + Saldeo + Power BI
I recommend this solution to companies with a turnover above PLN 10–20 million.
The ERP supplies:
receivables,
payables,
customer orders,
orders placed with suppliers,
stock levels.
Saldeo supplies:
current cost documents.
Power BI builds: 21-day cash flow
It shows the weeks, inflows, outflows, the balance of inflows and outflows and, for example:
the largest debtors,
the largest outgoing payments,
liquidity alerts.
This kind of solution fits the idea of building a modern finance function supported by systems and reporting tools.
Model, Level 3 – ERP + AI + rolling cash flow forecast
This is the direction that will dominate in the coming years. The system does not just show the future. The system starts to predict it.
An example:
A customer has a 30-day payment term.
AI analyses the history and sees that the customer:
pays after 47 days on average,
pays after 58 days during the summer holidays,
pays after 65 days when the amount exceeds PLN 50,000.
The forecast inflow is therefore calculated not according to the invoice but according to the customer’s actual behaviour. That is forecasting of an entirely different quality.
What architecture do I recommend for SMEs?
For most WK Profit Consulting clients, the following is enough:
Data sources
ERP
Saldeo
the bank
Integration layer
Power Automate
Make.com
Zapier
AI programs that integrate everything, such as Copernicus One for companies with a turnover above PLN 50 million.
Reporting
Excel
Power BI or similar
AI
Microsoft Copilot, ChatGPT and others.
an AI agent that analyses receivables
A solution like this does not cost hundreds of thousands of zlotys, yet every day it gives the owner an answer to these questions:
Will we run out of cash in three weeks’ time?
Which customer poses the greatest risk?
Which payments should be deferred?
Will a working capital loan be needed?
How will liquidity change once a new contract is won?
The biggest barrier is not technical
Having analysed more than 240 companies, I have noticed that the problem is usually not a lack of systems.
Most often the problem is the lack of a process.
Companies have:
an ERP,
a CRM,
an accounts department,
Saldeo,
online banking,
but nobody owns the cash flow process.
In my book and in my course, a great deal of space is devoted to responsibility for cash flows and to financial awareness at every level of the organisation.
My recommendation: first the process and responsibility for cash flow. Only then automate.
Automating chaos only leads to wrong forecasts being generated faster. A well-designed process, supported by ERP, Saldeo, Power BI and AI, on the other hand, makes it possible to monitor liquidity practically in real time and significantly reduces the risk of the company running out of cash.