Data Science Professional | Machine Learning Engineer | GenAI Engineer | Python Dev | Team-Lead

This initiative represents a cutting-edge data science project aimed at crafting a scalable, robust, and highly accurate demand estimation model for better resource and delivery planning. Harnessing the full potential of Azure Cloud's advanced capabilities, the project employs state-of-the-art machine learning algorithms, Sentiment Analysis via LLMs, Model-Explanations via Dash and GenAI-Agents, large-scale data processing pipelines and cloud-based infrastructure to forecast demand across all business units. By optimizing production and logistics, this model drives efficiency and elevate decision-making.

- Creation of an end-to-end workflow tailored to accurately predict demand patterns and behaviors for all business units
- Two-Pronged Approach for Intermittent Demand Forecasting
 ▪ Binary Prediction: Identifying the occurrence of the next demand instance.
 ▪ Regression: Estimating the magnitude of the demand.
- Top-Down Forecast pronged together with Bottom-Up-Forecast to increase performance
- Implementation ensemble techniques to bolster prediction precision.
- Use of Azure Synapse for streamlined scalability and accessibility.
- Leveraging Large Language Models (LLMs) and Azure OpenAI to implement sentiment analysis for customer reports, enhancing qualitative insights.
- Implementation of a GenAI-Agent-System to use data and model logs to explain model outcome to model-users
- Deployment of the GenAI-Agent-System as plotly-Dash-App via Docker and Kubernetes
- Deployment of MLFlow-Tracking-Server via Docker and Kubernetes

Development of a robust, scalable, and highly accurate car price estimation model using advanced data science techniques by leveraging the advanced capabilities of Google Cloud Platform (GCP) and VertexAI. Automization of time-consuming national-vehicle-code matching via fuzzy matching supported by LLMs (GPT-4o and Gemma from Google).
Proof of Concept for car-equipment-matching via LLMs to standardize the equipment overview on the website.
Proof of Concept for a car-suggestion Chat-Bot to help customers inform themselves about cars that fit their needs and suggest them fitting cars.

Implementation of a monthly Sales-forecast based on economic factors. The results reach the accuracy of the manual forecast are used for decision making to saving time. Knowledge sharing regarding deployment on Azure-ML or AWS Sagemaker.