I am a Malte, a Senior Data Scientist with a focus on computer vision and time series analysis. I am also very interested in MLOps and enjoy getting my hands dirty with code. I have started and ran my own business, so I have a lot of experience working on innovative projects. In addition to all this, I studied business administration in University, so I have a strong foundation in the basics of business.
Recent Projects
July 2022 - Present | Senior Data Scientist | Datadrivers GmbH
[5] Working on Recommender Systems in the AWS Cloud. Mainly focusing on the use of Amazon SageMaker and AWS Glue for postprocessing tasks.
May 2021 - June 2022 | Data Scientist | Datadrivers GmbH
[4] We have developed a temporal fusion transformer, which is used for multivariate time series analysis and demand quantity prediction. This project was a great experience for me, due to the size of the data and the fact how to create a data driven pipeline with PubSub, Dataflow and Tensorflow.
January 2020 - June 2022 | Data Scientist | Datadrivers GmbH
[3] In a recent project with 60,000 product descriptions that I have processed through a full spacy data pipeline, including GCP Cloud Run jobs. I have also developed an Annoy Index for similarity scores and processed these data to a Streamlit application.
[3] I have 40,000 images that I have processed through a full data pipeline, including GCP, Tensorflow, and Dataflow. Additionally to this setup I have used a pretrained MobileNetV2 and Streamlit applications as the WebUI.
[2] In this project, we have used a Thompson Sampler (Multi armed bandit) to automate the frequency of ads for a Google Cloud Platform customer. We built a model that predicts which ads frequency will be most effective for a given customer, then used Airflow to run the model on Google’s infrastructure. The end result was a significant increase in ad click-through rates and an overall improvement in customer satisfaction.
Before January 2020 | Data Analytics Consultant | Datadrivers GmbH
[1] During this time I deepened my knowledge in the field of neural networks. A wide variety of technologies and libraries were used to achieve this goal. Using GPUs, various problems could be solved and optimized using PyTorch and Tensorflow. Additionally I have created several Datawarehouse projects in the Cloud and on-prem. A larger selection of projects can be found in the Github repository.
