Case Studies Data Thinking

Selection, testing and implementation of a suitable ML mgmt. tool to professionalize ML lifecycles

A voice service provider wants to scale by integrating external software including additional speech understanding algorithms (AI) into its own voice chain.
This required, among other things, a solution that combines the two different, non-overlapping, speech understanding algorithms (AI-2-AI). This requires continuous end-to-end testing of the voice chain to continuously improve the integration.
mm1 has the project management and product ownership

A logistics service provider uses machine learning (ML) methods to optimize its core business, among other things.
Due to its organizational structure, ML applications are developed centrally and deployed decentrally. Hundreds of ML applications are currently productive, with several thousand models trained to date
The lifecycles of ML applications are not systematically supported. This leads to a huge manual effort in managing the lifecycles and selecting the best model for a given use case. Therefore, a tool shall be selected, tested and implemented according to user requirements