Who is it for?
This solution is suited for customers with business questions like the following:
- Will the customer by or not buy?
- What customers are at risk of churning?
- Is the production machine probable to break?
- A customer that bough product A is he probable to buy product B as well?
- What is the probability a new customer will become an active customer?
Answering these questions based on statistical prediction models is called machine learning.
Work methodology:
We work together with our customers to define the business requirement and the expected results from the process.
Following we plan a statistical model, a data panel and define the set of data required to test the statistical model.
We test the data set on a number of statistical algorithm alternatives. The most accurate algorithm was selected after comparing all the algorithm results with statistical tools.
We offer a monthly service to maintain the model and improve the results.
Our relative advantage:
Our relative advantage is manifested in a number of levels:
- Many years of experience in building a wide range of models.
- The use of advanced statistical methods for monitoring and calibrating the models.
- We have extensive technological knowledge and development capabilities that enable us to develop customized solutions to help integrate the model into the customers organization.
Examples for consumers clubs:
- Prediction of churn
- Next best call (Next best customer to call)
- Next best offer (Next best offer targeted to customer)
Examples of production lines:
- Reduce the number of faulty products produced.
- Prediction of production line machine break down.
Experience and clients:
We have extensive experience in developing models that have helped customers in making more informed decisions in a variety of business areas: health and medicine, online, finance, trading ,industry and more.