10 Fundamental Principles for Machine Learning Engineering

The hype around Machine Learning and Artificial Intelligence can give the illusion that software with integrated ML models is easy to develop. However, according to various studies, many ML projects fail to get into production. In this talk, Larysa will present 10 fundamental practices for machine learning engineering to succeed in your own project.

Transcript

8. Prediction Learning Domain Predictive Engine
30. Companies begin to explore ML/AI and its potential Short-term projects ML deployed at ad-hoc basis MLOps: Repository for ML models Manual deployment ML use cases are aligned with business objectives Pipelines for data processing and model training ML work fl ow for training, evaluation, batch prediction ML models are exposed through the API Utilizing pipelines Innovation by ML/AI Product-speci fi c AI-teams Establishing patterns and best practices Advanced MLOps: Feature stores, versioning, CI/CT/CD Fully automated processes