To be sure, an increasing number of companies are gathering initial experience with machine learning – from Proof of Concepts (PoCs) to Minimum Viable Products (MVPs). However, many of such projects are immature or are not being consistently followed through. Many potentials of the technology are thus unexploited. The reasons for this are many: High requirements of computing power, scalability of applications or inadequate monitoring of models.
Therefore, right from the start, a methodical procedure for the implementation of machine learning is desirable. This should include not only the technological aspects and algorithms, but also the organization and processes.
In our MHPDeepDive "Machine Learning in Production", we will first examine in detail why many projects are not being consistently followed through. Taking this as a starting point, we will present the CRISP AI methods developed by us, with which the various challenges can be met and the scaling of machine learning PoCs is successful. And: You will find out what the differences are between classic software applications and machine learning applications.
What awaits you in our MHPDeepDive
- We will show you why machine learning projects often involve more challenges than classic software development
- You will be provided with an overview of the methods developed by MHP for the introduction of machine learning in production
- You will find out how machine learning PoCs can be scaled successfully
Speaker:
Nikolaj Waller
Senior Consultant | Data Science & AI
