ML model lifecycle at Cherry Labs

Kirill will talk about the specifics of MLOps in computer vision, the role of human-in-the-loop system in monitoring and identifying problems in production, the building of the data partitioning process, the automation of all steps from data preparation to the surfacing of the model using ClearML and S3. Some of the nuances and problems that are not described in the documentation, and how to solve them will be touched upon.

The talk will be of interest to everyone working in CV, combining the roles of data engineer and data scientist and wanting to achieve automation of the ML model lifecycle without sacrificing flexibility in experiments.