Forecast Agent
Calculate rolling forecast - extrapolate trends, model scenarios, analyse sensitivities.
Prepares historical data, extrapolates trends, models best/base/worst scenarios, performs sensitivity analyses and compares the forecast with the budget.
Score Dashboard
What This Agent Does
The rolling forecast replaces the rigid annual plan with a continuously updated projection. Every month or quarter, the forecast is recalculated - with updated data and adjusted assumptions. The challenge: the mechanical work (data preparation, trend extrapolation, scenario calculation) consumes the time the controller needs for strategic assessment.
The Decision Layer clearly separates: preparing historical data, extrapolating linear trends, calculating scenarios and analysing sensitivities - these are automatable tasks. Defining assumptions, plausibilising results and assessing strategic options - that stays with the controller.
The result: the forecast is ready by day 3 after month-end instead of day 10. Three scenarios are fully calculated. The sensitivity analysis shows which variable has the greatest impact. And the controller can use their time for strategic interpretation.
Micro-Decision Table
Prepare historical data Which data forms the basis for the forecast? Rules Engine
Database query and data cleansing
Decision Record
Challengeable: Yes - rule application verifiable. Objection possible for incorrect data or wrong rule version.
Calculate trend extrapolation How do metrics develop if the trend continues? Rules Engine
Linear trends = R, non-linear patterns = A
Decision Record
Challengeable: Yes - rule application verifiable. Objection possible for incorrect data or wrong rule version.
Define assumptions Which growth, cost and FX assumptions apply per scenario? Human
Strategic assumptions require human judgement
Decision Record
Challengeable: Yes - via manager, works council, or formal objection process.
Calculate scenarios What do best, base and worst case yield? Rules Engine
Arithmetic calculation per scenario
Decision Record
Challengeable: Yes - rule application verifiable. Objection possible for incorrect data or wrong rule version.
Sensitivity analysis Which variable has the greatest impact on the result? Rules Engine
Calculation = R, interpretation of relationships = A
Decision Record
Challengeable: Yes - rule application verifiable. Objection possible for incorrect data or wrong rule version.
Compare forecast vs. budget How does the forecast deviate from the budget? Rules Engine
Arithmetic comparison
Decision Record
Challengeable: Yes - rule application verifiable. Objection possible for incorrect data or wrong rule version.
Plausibility check Are the forecast results plausible? Human
Experience and industry knowledge
Decision Record
Challengeable: Yes - via manager, works council, or formal objection process.
Strategic assessment Which strategic options emerge from the forecast? Human
Judgement in strategic interpretation
Decision Record
Challengeable: Yes - via manager, works council, or formal objection process.
Decision Record and Right to Challenge
Every decision this agent makes or prepares is documented in a complete decision record. Affected parties (employees, suppliers, auditors) can review, understand, and challenge every individual decision.
Prerequisites
- ERP system with at least 24 months of historical financial data
- Planning system for budget data (SAP BPC, Jedox, Anaplan or equivalent)
- Defined scenario parameters and assumptions framework
- Access to external data for market and FX assumptions
Governance Notes
Not GoBD-relevant: the Forecast Agent does not process tax-relevant data - it creates projections and scenarios. However, the results can have significant impact on investment, financing and personnel decisions.
The assumptions and strategic assessment remain with the controller. The agent delivers the calculation - responsibility for assumptions and recommendations lies with the human. This must be transparently documented.
§203 StGB-relevant data is encrypted end-to-end and never passed to AI models in plain text.
Process Documentation Contribution
Infrastructure Contribution
The Forecast Agent uses the variance analysis engine of the Budget Variance Analysis Agent and the KPI library of the Management Reporting Agent. The scenario modelling framework becomes the standard for all planning and projection processes. The sensitivity analysis engine is reused by the Cash Forecasting Agent.
Builds Decision Logging and Audit Trail used by the Decision Layer for traceability and challengeability of every decision.
Related Agents
Budget Variance Analysis Agent
Calculate plan/actual variances, identify causes, create commentary draft.
Management Reporting Agent
Calculate KPIs, analyse trends, detect anomalies, prepare board reporting.
Frequently Asked Questions
How often should the rolling forecast be updated?
Most companies update monthly or quarterly. The agent supports both rhythms. A monthly forecast is more accurate but requires more effort in maintaining assumptions. The choice depends on the volatility of the business model.
Can the agent also incorporate non-financial drivers into the forecast?
Yes, if data is available. Sales volumes, headcount, production capacity and other drivers can be included as variables in scenario modelling. Defining the relationships between drivers and financial metrics is a one-time configuration.
How is forecast accuracy measured?
The agent automatically calculates forecast accuracy by comparing past forecasts with actual results. Systematic analysis of forecast errors improves model quality over time.
Implement This Agent?
We assess your finance process landscape and show how this agent fits your infrastructure.