From a model to production like a Pro: Software-engineering best-practices


In this fast-paced, pragmatic, tips-oriented and structured talk, Marcel Krčah will cover all software engineering aspects required for industrializing and running a data-science / machine-learning / big-data model in production: team collaboration, codebase, architecture, dev-ops, data-pipelines, ETL. Applying those principles helped Marcel and his teams to deliver data-science products rapidly, with low-time-market and low-maintenance.

The talk is suitable for anyone responsible for running a data-science / big-data product in production.

Marcel is a senior data engineer and project lead. Founder of DataSharks, a solo data engineering consultancy in Netherlands, he helps companies to kick-off big-data projects and to industrialize data-science proof-of-concepts, with the use of modern tech stacks. He led an engineering team in AirFrance/KLM to kick-off Prognos - a strategically important platform for predictive airline maintenance. Currently, he leads engineering efforts within KLM Operations Research, helping a team of 30 data scientists to reach faster time-to-market while minimising future product maintenance. Apart from project work, Marcel is a speaker and a trainer in the field of (big) data engineering.