ML-based Systems at Deployment

A focused symposium course on the deployment of Machine Learning (ML) models as part of larger software systems. We explore the ML context and definition, the motivation and challenges behind deploying ML models, the data science methodology, the ML pipeline, and current software architectures for ML-based systems, enabling participants to understand and implement effective ML deployment strategies.

Department: 34th International Symposium on Statistics

Institution: Universidad de Nariño

Lectures: 3

Hours: 4.5

From: July 30, 2025

To: August 01, 2025

Lectures

ML-based Systems

July 30, 2025

This lecture presents the Machine Learning context, its current narrative, and the motivation and definition of the ML-based Systems concept. We start this short course by developing a couple of examples to motivate a systems perspective when designing, developing, and deploying systems based on ML.

ML Adoption

July 31, 2025

This lecture will start looking into the adoption process of ML technologies. The goal is to design ML-based systems that align with our socio-technical systems and their stakeholders, while ensuring its careful development and safe deployment. To this end, we introduce simple but powerful methodologies for designing and developing ML-based systems.

ML Deployment

July 30, 2025

This lecture will start looking into the data dimension of the ML concept and will emphasise on the importance of data-orientation. We first define the concept of data, the associated challenges, and provide examples of data collection processes. We then define a data science methodology to iteratively build the datasets that will feed our machine learning models. This lecture explores the first step of this methodology, "data access".