- Continued building forecasting model for commercial use within Etos. Goal at this stage is to set up an MVP that translates improved accuracy from POC into a production-ready forecasting process- Introduced MLOps principles into the forecasting process: unit & integration testing, CI/CD pipeline, data quality monitoring and model registry - Solution ready for production for financial planning & budgeting purposes

- Started building from scratch a forecasting model for commercial use within Etos. The use cases of the model are financial planning and promo demand planning. Goal in this stage is a POC showing improved forecast accuracy compared to business forecasts.- Used Databricks, Azure, MLFlow

- Set up and lead data science startup with products in operations and marketing area- Implemented transport planning solution for clients in several industries that showed demonstrably improved routing compared to old plan- Used state-of-the-art data engineering to make solution scalable: Docker, Docker Swarm, CLI and AWS

- Acted as technical expert and lead analyst on forecasting solution for global FMCG & pharmaceutical companies. The solution was capable of predicting demand for newly introduced products (NPDs) based on customer survey and historical demand data- Presented on conjoint analysis at 2020 Sawtooth Software European Conference

- Played a pivotal role in the start of an Advanced Analytics team that manages data science solutions based on questionnaire data- Executed with limited oversight several marketing research project using conjoint analysis

- Presented theory and assignment solutions to groups of approx. 30 students- Guided students with assignments- Checked assignments and examsCourses: BSc 2 - Combinatorial Optimisation (2016-2017)BSc 2 - Econometrics 1 (2016-2017 and 2017-2018)BSc 2 - Simulation (2016-2017)