AI Agents in Corporate Planning

Corporate planning is becoming more complex. Markets are changing faster, cost structures are becoming more volatile, and new influencing factors must be taken into account with ever-shorter lead times. At the same time, business units expect reliable forecasts and quick decisions. In many planning projects, however, a familiar pattern emerges: forecasts are maintained in different systems, Excel files are exchanged, and planning assumptions are reconciled manually. The effort increases, while transparency decreases. Modern planning solutions, therefore, take a different approach. Instead of constantly providing new reports and dashboards, the focus is on direct interaction with data, forecasts, and planning models.

This is exactly where exciting developments are currently taking place around SAP Analytics Cloud, SAP Business Data Cloud, Joule and AI Agents. Our observations from customer projects are clear: The goal of planning remains unchanged. What is changing is the way people work with planning. The shift is moving from searching for information to engaging in dialogue with data. What’s crucial here is not just interaction with data, but the ability to intelligently incorporate individual business logic and forecasting models. This is precisely where new approaches involving AI agents and modern forecasting solutions are currently emerging.

Many companies face the same challenge: the number of relevant factors is constantly growing. These include, among others:

• Inflation
• Energy Prices
• Volatile commodity markets
• Geopolitical developments
• Skilled Labor Shortage
• Fluctuating demand

Traditional planning cycles are increasingly reaching their limits. Forecasts are created, coordinated, and approved. However, the operating conditions often change even while the process is underway. That is why continuous business planning is becoming increasingly important. Decisions must be made more quickly, and forecasts must be updated more frequently. The speed required to do this is difficult to achieve with manual processes alone.
 

In nearly every AI project, the same lesson emerges: The quality of the results depends directly on the quality of the data. A poor data set leads to poor AI. A good data set enables reliable predictions and scalable AI use cases. For this reason, the real key to success often lies less in the model than in the data set. Data must be consistent, quality-assured, domain-specific, and reusable. This is precisely why data products and data democratization are becoming increasingly important. Instead of isolated data silos, domain-defined information building blocks are emerging that can be shared across different departments.

SAP addresses this approach with the Business Data Cloud. The focus is on qualified data products that provide not only the actual data but also the business context. Key metrics, business rules, and relationships are already described and made available in a standardized format. The benefits of corporate planning are significant. AI no longer works with isolated raw data but rather with information understood in a business context. This improves data quality, governance, transparency, reusability, and the traceability of analyses. Only on this basis can intelligent planning processes be established in a sustainable manner.

When it comes to AI, the question often arises as to how individual forecasting solutions and SAP Joule relate to one another. The answer is simple: They have different objectives. Joule focuses on interacting with data, analytics, and applications. SAP is continuously developing Joule as an analytical assistant and is making strategic investments in agent-based scenarios.

Our forecasting solution, on the other hand, focuses on:

• Forecast Quality

• Customized forecasting models

• Individual factors

• Technical forecasting logic

The two approaches complement each other perfectly. While Joule enables interaction, the forecast models provide the technical foundation for future agent-based planning scenarios.

Planning is becoming increasingly interactive. Instead of working through reports, key metrics, and planning templates, business units will be able to interact directly with their planning models. For example, business units could ask:

"What happens to the total cost if we plan for five additional positions?"

An intelligent agent understands the request, takes existing planning contexts into account, uses forecast models, and generates a new forecast. The key added value doesn’t come solely from the chat window. The real added value comes from the agent’s ability to link data, business logic, and forecast models.

AI agents go far beyond traditional chatbots. They can analyze information, assess impacts, derive possible courses of action, and coordinate multiple process steps. This concept becomes particularly interesting in the context of integrated business planning. Sales planning, workforce planning, production planning, cost center planning, and financial planning are interconnected through shared relationships. Modern enterprise planning approaches create a seamless flow of data for this purpose and reduce the need for manual reconciliation. It is precisely in this context that AI agents will be able to provide support in the future by automatically identifying impacts and evaluating them from a business perspective.