.svg)
Organizations have wanted to be able to predict the future better for decades. Whether it's revenue forecasting, demand forecasting, churn analysis or financial planning, better forecasts lead to better decisions. Traditionally, this challenge has been addressed with predictive models and machine learning. These techniques are powerful, but they also have clear limitations. With the rise of generative AI and so-called reasoning agents a new approach is now being created that could potentially have a major impact on how organizations make predictions.
Predictive models based on machine learning have become extremely popular in recent years. They can recognize patterns in historical data and predict future outcomes based on that.
But these models have a number of distinct features:
1. Large amounts of data are required
To make reliable predictions, machine learning models often require extensive data sets. Not only over a long period of time, but also with many different features. Examples include customer behavior, economic indicators, marketing activities or seasonal effects.
2. High development costs
Building a good predictive model is labour-intensive. Data scientists must collect data, clean it, select features, train and optimize models. This process can take weeks or even months.
3. Customization per use case
Many models are specifically designed for one problem. For example, a churn model does not automatically work for revenue forecasts. As a result, organizations often have to develop and maintain multiple models.
All of this makes traditional predictive AI powerful, but also expensive and complex.
With the breakthrough of generative AI and large language models, we are seeing a new category of AI systems emerge: reasoning agents.
These systems are not primarily built to learn statistical patterns from huge data sets. Instead, they have been trained to to reason logically, to calculate scenarios and to make connections based on available information.
Thanks to massive advances in model architectures and computing power, these agents are able to:
This raises an interesting question: can reasoning agents still make reliable predictions with relatively little data?
To investigate this, a practical case was carried out at an organization that wanted to make financial forecasts. In this case, two approaches were put side by side:
Both systems received the same business context and comparable input data.
The result was surprising.
The reasoning agent's forecasts were approximately 90% in line with the machine learning model predictions.
In other words, despite significantly less data preparation and model training, the reasoning agent was able to generate almost the same results.
This suggests that logical reasoning combined with limited data can be surprisingly powerful in many situations.
This development has a number of important implications.
Because reasoning agents rely more on logic and context, they can often work with less historical data.
This is particularly interesting for organizations that:
A traditional machine learning process can take months. Reasoning agents can often within days or weeks are deployed.
Because less specialized model development is required, predictive AI is becoming more accessible to a larger group of organizations.
While reasoning agents have a lot of potential, there are also important caveats.
In machine learning, the biggest investment is often in the development phase: collecting data, training models and optimizing them.
With reasoning agents, this cost structure shifts.
The model itself has already been trained, but every time you use it, it must thinking, analyzing and calculating. This is done via tokens and compute, which can result in considerable costs if used intensively.
A well-trained machine learning model often produces very stable results.
Reasoning agents can be more variable because they perform each analysis over and over again.
In statistical models, it is often clear which variables influence the result. With reasoning agents, it can be more difficult to fully control that influence.
We are at an interesting tipping point.
Traditional machine learning models remain extremely valuable, especially in situations where:
But reasoning agents offer a faster, more flexible and more accessible route to predictive insights.
In many cases, organizations are likely to opt for a hybrid approach:
The development of reasoning agents shows that predictive AI is no longer solely dependent on complex machine learning pipelines and massive data sets.
Experiments show that this new generation of AI systems up to 90% can achieve comparable results in forecasting, with significantly fewer data requirements and development time.
That makes predictive AI more accessible than ever.
At the same time, this new approach requires a different way of looking at costs, governance and implementation.
What's clear: reasoning agents are getting closer to the power of traditional predictive models — and that can fundamentally change how organizations make decisions.
.svg)
