Marcel Steger-Senior AI Engineer

Project Tech Stack: Python, AWS, Azure, FastAPI, openai, pandas, unittest/pymock

Achievements:

• Engineered automated data extraction pipelines to transform complex Excel datasets into structured formats via LLM-driven workflows.
• Architected a generative slide-deck engine that translates natural language prompts into formatted presentation assets.
• Integrated advanced LLM capabilities with the OpenAI Response API, implementing sophisticated tool-calling and structured output logic.
• Developed and containerized scalable backend microservice using FastAPI, Docker, and OpenShift to host and serve agentic skills.

Project Tech Stack: Agentic AI, FastMCP, Kafka, unique-sdk, Agents, openshift

Achievements:

• Architected an autonomous agentic system for insurance document extraction using advanced LLM orchestration.
• Developed a custom MCP (Model Context Protocol) Server to bridge internal backend tools with LLM-driven interfaces, enabling seamless tool-use and agentic autonomy.
• Engineered highly reliable data extraction pipelines using OpenAI Structured Outputs, ensuring 100% schema adherence via Pydantic.
• Built automated business process workflows (Agentic Tool Use) for complex invoice and contract processing.

AI Engineering, Data Science, and Software development

Project Tech Stack: Transformers, mflow, CUDA, Python, Pytorch, Label Studio

Achievements:

• Comparative tests of state-of-the-art Computer Vision AI models
• Extensive model testing
• GPU-accelerated model training
• Research and conceptualization of CNN architectures
• Responsibility for image processing pipelines and infrastructure
• Development of an end-to-end computer vision solution in Python
• Agile working methods

Data Science and Data Engineering

Project Tech Stack: Python, AWS RDS, AWS MWAA, Airflow, PostgreSQL, dbt, GitHub, Plotly & Dash

Achievements:

• Responsibility for data pipelines with Airflow MWAA
• Application code development for data dashboards
• Modularization of the existing codebase
• Dashboard development
• Time-critical deliveries to customers
• CI/CD pipelines
• Agile working methods

Project Tech Stack: GCP Compute Engine, Docker, BigQuery, Google Data Studio, GCP IAM, AWS Sagemaker, AWS VPC, AWS RDS, OpenVPN, PostgreSQL, JavaScript

Achievements:

• Creation and maintenance of a cross-cloud data pipeline for Bluetooth device communication data
• Development, delivery, and deployment of the solution in a Docker image
• Creation of a new Google Looker Studio community visualization tool in JavaScript: Sankey Diagram
• Customer behaviour analysis for marketing recommendations
• Algorithm parallelization
• CI/CD with Jenkins

Project Tech Stack: OpenCV, Python, Flask, Linux, RaspberryPi, CI/CD, Docker, PyTorch, scipy

Achievements:

• Development of a stable computer vision solution from R&D to go-to-market
• Implementation of a robust, efficient, and reliable algorithm for determining the optimal affine transformation matrix
• Implementation of a fast and reliable image segmentation algorithm
• Determination of optimal lighting conditions for image acquisition
• Development of an extensive test suite for quality control
• Management of the CI/CD pipeline
• Object-oriented programming throughout the project
• Post-launch customer support on-site for image processing system and network communication
• Agile working methods

Feature Selection Methods for Big Data

Project Tech Stack: R, Python, PySpark, Bash

Achievements:

• Successful implementation of a Python module for Big Data feature selection in PySpark
• Documentation of implemented functions and creation of a user manual
• Thorough research and understanding of feature selection methods
• Data restructuring, processing, and preparation, exploratory data analysis
• Code parallelization in R