Make the input easy to give and hard to get wrong. Ask each budget owner a few fixed questions in one format, pre-fill what your ERP and bank data already show, and let AI flag answers that do not fit the history. A treasury or FP&A lead reviews the inputs and owns the forecast that goes to the CFO.
#Key takeaways
- Cash forecasts usually fail on inputs: late, inconsistent or optimistic numbers from sales, procurement and projects.
- AI helps on collection, not prophecy: pre-filling from receivables, payables and orders, chasing missing inputs and flagging outliers.
- The forecast model stays in your spreadsheet or treasury tool, and a person owns the number.
- Business units give better inputs when they see the forecast and the variance from their own last input.
- A shared template and a fixed submission day fix a lot before any AI is involved.
#What does a cash forecasting cycle look like for treasury?
Every week the treasury analyst builds the rolling cash view. Receipts come from the aging report, payments from the AP run and payroll, and the rest from emails to sales, procurement and project managers asking what they expect to collect or spend. Half reply late, some send last week's file, and one big customer receipt turns out to be a hope.
This is a well-known pattern. euromoney, checked november 2026 reported that multinationals found it hard to get business units to understand why cash forecasting matters, and the AFP FP&A benchmarking survey, 2025 found that about six in ten respondents named data reliability, and as many named data accessibility, as a challenge.
#Where does AI help with cash forecast inputs?
- Pre-fill each owner's input from what you already know: open receivables, approved purchase orders, scheduled payroll, loan repayments.
- Send each owner only the questions that need their judgement, such as which large receipts are at risk or which projects slip.
- Compare each answer with history, for example a customer that always pays late forecast to pay on the due date, and flag it to the analyst.
- Read replies in email or Teams, in Arabic or English, and load them into the template.
- Draft a short note on the biggest movements for the treasury lead to edit.
Keep with people: the assumptions, the funding decisions, the forecast sent to the CFO or the bank, and any judgement on covenants. AI prepares inputs; it does not decide the cash position.
#What does a practical playbook look like?
- Agree a fixed submission day and one template for every owner.
- Ask fewer questions: large receipts, large payments, and changes since last time.
- Pre-fill everything the ledger already knows, so owners confirm rather than type.
- Send each owner their own forecast against actuals, so accuracy becomes visible.
- Measure inputs on time, forecast against actual by category, and the number of late corrections.
The AFP, checked november 2026 makes a related point in its best practices for cash forecasting: treasury relies on business units for input, and should send the key forecast information back so those units can act on it.
An illustrative example, the numbers are assumptions: a group with eight budget owners might find that two of them account for most of the forecast error, which tells you where to focus first.
#When is a template and a weekly call enough?
If you have one entity, a few large customers and a treasury analyst who knows every receipt, a good template, a fixed day and a short weekly call are enough. Your bank or treasury tool may already pull balances automatically. AI earns its place with many owners, several entities and banks, or inputs that arrive in mixed formats and languages.
#How does Stennir approach cash forecasting?
Stennir starts with AI consulting, because the first question is which inputs drive the error. The AI Opportunity Audit returns five ranked workflows, each with a cost and a payback case, and one recommended build with its scope and hosting region. The report is yours whether or not you build with Stennir.
If input collection is the build, AI implementation connects it to your ERP, aging report and template, with a treasury lead approving the inputs and every action logged. AI training helps budget owners and analysts work with it.
We are
Stennir improves the inputs that feed your forecast, using your own data, with a treasury lead owning the number.
We aren't
Stennir is not a treasury management system, and does not produce a cash forecast that nobody on your side signs.
#What should a treasury lead do next?
Compare your last few forecasts with actuals by owner and category, and find which inputs moved the most. That is your first workflow. Read more at AI for finance and how to pick the first process to automate with AI, then take the free readiness assessment or book a 30-minute call.
#Common questions
Can AI predict our cash flow on its own?
It can find patterns in payment history, such as which customers pay late, and suggest an adjustment. It should not replace the inputs from people who know about a large contract or a delayed project. The forecast stays in your model, and a treasury lead decides which suggestions to accept.
What do you need from us to start?
Your current forecast template, a few past forecasts with actuals, exports of receivables and payables, and the list of budget owners who submit inputs. The treasury or FP&A lead who builds the forecast today explains the steps and is named as the approver.
Does it work across entities and banks?
Yes. Inputs can be collected per entity and per bank account, then rolled up in your template. Owners can reply in Arabic or English. Each entity's access can be limited to its own data, and the group view stays with the treasury lead.
Where does our cash data stay?
In your own accounts and the hosting region your rules require. Stennir does not copy bank balances or ledgers into its own systems. Access is granted by you, every action is logged, and you keep the accounts, the data and the off switch when the work ends.
