06.09.2022 aktualisiert

**** ******** ****
100 % verfügbar

Data Analyst

Frankfurt am Main, Deutschland
Deutschland
Materials Engineering
Frankfurt am Main, Deutschland
Deutschland
Materials Engineering

Profilanlagen

Lebenslauf
Software-Projekte

Skills

ABOUT ME

I’m a freelance Data Analyst active in the metallurgical industry in Europe.
I support industrial companies on their road to digitalization, step by step and always with the business in mind.
My metallurgical education and industrial experience enable me to understand each problem on a technological level. 

Languages: German, English, Russian, Dutch, French

SERVICES
  • Data Analysis and Machine Learning
  • Data Engineering (ETL)
  • Teaching and Consulting
TECHNOLOGIES
  • Programming: R,  Python, Matlab & Simulink
  • Machine Learning: scikit-learn, PyTorch, Keras-TensorFlow
  • Tools: Apache Airflow, Shiny, Git, Tesseract OCR
  • Deployment: Docker, Google Cloud Platform
  • Databases: SQL Server, MySQL, PostgreSQL, Cassandra
EXPERIENCE

Junior Consultant Data Science (full-time)
STATWORX GmbH
Mar 2020 – Aug 2020
Frankfurt am Main, Hesse
  • Development and deployment of transformer-based machine learning models in Python for Natural Language Processing on Google Cloud Platform.
  • Improvement of Optical Character Recognition performance of Tesseract OCR on German invoice documents.

Process Engineer (full-time)
TRIMET Aluminium SE
Feb 2018 – Jan 2020
Essen Area, Germany

  • Developed a dashboard in R and Shiny to evaluate process control performance across two plants in Germany.
  • Developed a dashboard in R and Shiny to estimate the tapped metal mass from individual ovens based on a tapping vehicle’s total mass.
  • Developed two process control modules in C through model based design with Matlab, Simulink and Stateflow.

Erfahrung

Berufserfahrung0 - 5 Jahre

Sprachen

DeutschMutterspracheEnglischgutFranzösischgutNiederländischgutRussischgut

Projekthistorie

Data Engineering: ETL-development with Apache Airflow

Not disclosed

Transport und Logistik

10-50 Mitarbeiter

Development of an ETL-pipeline as a microservice:
  • Transfer data from the Cassandra IoT-database to an MySQL application database
  • Deployment in a Docker container

Personal Workstation for GPU-Computing

Hobby project
PERSONAL WORKSTATION FOR GPU-COMPUTING
  • Researched and assembled all hardware components
  • Installed and configured the Arch Linux operating system

Cognite Open Industrial Data Project

Hobby project

Industrie und Maschinenbau

< 10 Mitarbeiter

Designing an automated data processing and visualization workflow to monitor real sensor data from an industrial compressor
  • Scheduled a data pipeline in Python and Apache Airflow that writes a daily batch of sensor data to a PostgreSQL database
  • Developed a Shiny Dashboard showing pressure difference and sensor location
  • Deployed the project using Docker on a linux virtual machine on Google Cloud Platform: http://34.89.171.53:3838/

Dashboard for Gas-Treatment time series Visualization

not disclosed

Industrie und Maschinenbau

50-250 Mitarbeiter

Developed an automated workflow and dashboard to process and view the KPI’s of three gas-filtering modules
  • Automated R scripts that process a daily batch of data from raw text files
  • Developed a Shiny dashboard that can be customized by a user with no coding experience via an Excel sheet

Machine Learning PoC for Natural Language Processing

not disclosed

Internet und Informationstechnologie

10-50 Mitarbeiter

Co-developed a proof of concept for a large German IT-group to extract and classify accounting information from German invoices
  • Developed a workflow in Python and Bash to evaluate optical character recognition performance
  • Experimented with multiple PyTorch transformer-models for NLP to classify invoice line items into three budget categories
  • Retrained the final model in TensorFlow and deployed it using Docker on Google Cloud Platform as a microservice

Dashboard for Metal Production Visualization

not disclosed

Industrie und Maschinenbau

250-500 Mitarbeiter

Developed a Shiny dashboard mounted on the SQL Server production database that calculates the metal production of 360 units.

Computer Vision Competition

Hobby project

Industrie und Maschinenbau

< 10 Mitarbeiter

Developed a PyTorch model for the Steel Defect Detection competition on Kaggle whose test prediction finished in the top 23%

Anode-Changing Process Control Module

not disclosed

Internet und Informationstechnologie

250-500 Mitarbeiter

Developed a software module with MATLAB, Simulink and Stateflow to control the anode-changing operation
  • Documented the software design and usage instructions
  • Designed automated tests as part of a CI/CD pipeline
  • Tested the software module on two electrolysis pots on two different plants

Kontaktanfrage

Einloggen & anfragen.

Das Kontaktformular ist nur für eingeloggte Nutzer verfügbar.

RegistrierenAnmelden