IT Consultant

• Conducting IT consulting projects in data science and data management for various client companies in Germany and Switzerland.

• Completed client projects:

• Project at EUPD Research Sustainable Management GmbH (Bonn) (06/2024 – 12/18/2024):

• Analyzed a large number of sustainability reports using RAG (Retrieval Augmented Generation), Milvus vector databases and large language models (LLMs) provided via Watsonx.ai

• Tested prompts, LLM model types and parameters for answer quality and to avoid hallucinations

• Deployed analysis scripts in the cloud via Docker

• Developed a Streamlit app to generate answers in a user-friendly browser interface

• Used AI language agents to search for company information online to pre-classify sustainability reports by industry and number of employees

• Technologies used: Watsonx.ai, Python, Jupyter Notebooks, Docker, Streamlit, VS Code, GitLab

• Project at IGS Informatikgesellschaft für Sozialversicherungen GmbH (St. Gallen, Switzerland) (09/2023 – 12/18/2024):

• Extended Python modules to transform, store and import social security data into an online database system

• Mapped table structures using Python classes (column names, data types, field lengths, foreign keys, unique constraints, references to other tables)

• Automated extraction of data from Excel tables, created JSON files for intermediate storage, imported JSON data into DB2 database environments via batch scripts

• Logged SQL merge queries via the Python SQLAlchemy package for reuse in different database schemas

• Technologies used: Python, SQLAlchemy, DB2, VS Code, GitLab, DBeaver

• Project at Henkel AG & Co. KGaA (Düsseldorf) (09/2022 – 12/2022):

• Trained and optimized machine learning models in Azure Databricks to predict whiteness values measured by a stain monitor in washing experiments

• Created charts and visualized metrics to measure prediction quality in regression algorithms (including neural networks and random forests)

• Explained predictions using LIME and SHAP values

• Technologies used: Python, Databricks, MS Azure (as user)

• Independently managed a startup offering a solution for digital contact tracing via QR codes
• Carried out data-driven customer projects

• Topic: “Targeted preparation and sample analysis of sensor data to generate alerts using predictive maintenance for clients of a mid-sized family business”
• Objectives:
• Analyze IoT sensor data automatically uploaded to the SAP Cloud Platform from an online water analysis system
• Automatically send alerts using machine learning models in SAP Predictive Maintenance & Service Cloud Edition

Deployed by RPS Engineering GmbH.