Turn your data into measurable business value
We make your data speak with AI—through forecasts, real-time dashboards and anomaly detection. So you can make decisions based on facts rather than gut feeling.
What we do in practice
Prerequisite: clean data
We start realistically: Before models are trained, the data sources need to be assessed. In many SMEs, data is spread across five systems, three sprawling Excel sheets and the heads of two employees. We help you turn it into a format ready for analysis—without requiring a massive data warehouse project.
Models we typically use
- Conventional machine learning: XGBoost, LightGBM, Random Forest—powerful, explainable and resource-efficient
- Time series: Prophet, ARIMA, neural networks (TFT, N-BEATS)
- Deep learning where it delivers value (image data, text, sequences)
- LLMs to prepare and explain the results
Data privacy
To us, data analytics does not mean moving all data into a US cloud. We use pseudonymization, aggregation and—where possible—local processing. Personal data only leaves your area of control if you explicitly authorize it.
Frequently asked questions
What is predictive analytics?
Predictive analytics uses historical data and AI models to predict future events—sales, maintenance, churn and liquidity.
How much data do we need?
For seasonal forecasts, 18-24 months of historical data; for classification, 1,000-10,000 examples; and for anomaly detection, a few weeks of clean data are often sufficient.
Where are the results delivered?
Power BI, Tableau, Looker, Grafana, or embedded in your ERP or CRM. Also via webhook or API.
How accurate are the predictions?
That depends on the use case. We validate through backtesting and communicate confidence intervals.
How much does it cost?
A pilot typically costs 8-20 thousand euros, depending on the data situation. Ongoing model maintenance is offered for a monthly flat fee.
Which question should your data answer?
We will assess free of charge whether your data is already sufficient—or what needs to happen first.