SESSION

Efficient MLOps: Developing and Deploying ML Models with Databricks (repeat)

OVERVIEW

EXPERIENCEIn Person
TYPEBreakout
TRACKData Science and Machine Learning
INDUSTRYEnergy and Utilities, Enterprise Technology, Financial Services
TECHNOLOGIESAI/Machine Learning, GenAI/LLMs, MLFlow
SKILL LEVELIntermediate
DURATION40 min

This session is repeated.

 

This session explores how to design and implement an efficient MLOps framework to successfully ensure the development and deployment of ML models. The discussion embraces the management of diverse environments and the implementation of robust CI/CD pipelines. Emphasis is placed on standardized coding practices and presenting tools to ensure code quality and test coverage. The strategic use of Unity Catalog, MLflow, and MLOps stacks as crucial tools to manage and monitor the evolution of models and facilitate their progression through the various stages is discussed. Finally, advanced techniques such as Delta Live Tables and Lakehouse Monitoring are explored for their role in enabling robust, intuitive monitoring of input and output data drift to ensure model consistency. The creation of this framework aims to provide a standardized approach to developing ML models in all the countries in which Plentitude operates. To underscore its potential, a demonstration will be shown.

SESSION SPEAKERS

Lavinia Guadagnolo

/Product Owner - Data Scientist
Plenitude

Alessandro Mazzullo

/Data Scientist
Plenitude