Decision data scientist

Nightingale Health's core tech is a ML model that maps the NMR spectrum of a blood sample to biomarker concentrations. Its development is subjected to a strict Quality Management System, and the results are also validated against external laboratories. As part of the data science team I specialized in understanding that aspect of it. I was e.g. charged with coming up with an auditable way to show when the model had changed only insignificantly and the costly validation process could be skipped.

R, Airflow, Docker, PostgreSQL

IIoT project, 9/2018-7/2019 This project aimed to create a system for anomaly detection for predictive maintenance purposes, as well as reduce alert fatigue due to an overwhelming amount of signals already present. I was mainly responsible for understanding the processes the data described, and developing the anomaly detection model.
Tech: R, PostgreSQL, Power BI 

Sales forecast in FMCG, 3/2019-6/2019 A major fast-moving consumer goods manufacturer wanted a sales forecast in a foreign market in order to inform their strategic decisions there in the next few years. I was responsible for defining and fitting the model.

Tech: SPSS Modeler

In addition to various projects, I was responsible together with another senior colleague for the competence building of my more junior colleagues in biweekly workshops, on e.g. software development practices.

CSC is a non-profit state enterprise with special tasks. I was the maintainer of distributed R on the supercluster, and the customer support for its use both from the technical and the data science point of view. I was also a teacher of data skills to customers and colleagues (courses, book seminars, consulting, supervision of trainees), and a facilitator between customers and developers, particularly for the then emerging field of computational social science.
Tech: R, MPI, SLURM