1. Historical data
Use various historical customer data available in your company to create a data set. This dataset will be the input for predictive modelling.

2. Multiple format
The platform accepts multiple file formats, like MS Excel and character separated files (.csv).

3. Secure
Your data is safe with us. ClearPredictions is hosted on the Microsoft Azure platform (North Europe) in a private cloud. We use state of art security technology and are happy to sign a confidentiality agreement.

1. Data preparation
Your historical data will be automatically prepared and qualified for seamless processing in the various predictive methods we use.

2. Target setting
You determine which variable in your historical dataset will be the target for predictions. Typical predictions targets are churn, cross & upsell transactions, bad debt.

3. Data patterns
The platform automatically shows the influence of data towards the target variable. These insights might motivate you to adjust your internal company processes.

1. Fully automated
You simply push the button and ClearPredictions.com automatically builds a predictive model. By using a cross validation mechanism you immediately know the accuracy of the predictive model.

2. Machine learning algorithms
We use ten different machine-learning algorithms to create predictive models. Each algorithm has its own characteristics. So for all kind of datasets, whether big or small, having missing values in it or cleansed, an accurate predictive model will be built.

3. Scientifically proven
The machine learning algorithms we use are all scientifically proven and co-created with universities and top data scientists worldwide.

1. Actual data
By providing new customer data you are able to create predictions for your customers. The platform adds the outcome of the predictions to your actual data sheet. You can start making your predictions actionable.

2. Batch driven
By uploading a dataset containing new customer data you are able to make predictions for these customers.

3. Event driven
Your predictive model can also be used in real-time and in event driven scenarios. Examples of typical real-time environments are customer contacts centers and self service websites.