Profilbild von Markus Semmler Senior • Deep Learning • Probabilistic Graphical Models • Reinforcement Learning • AI-Software aus Darmstadt

Markus Semmler

verfügbar

Letztes Update: 20.03.2024

Senior • Deep Learning • Probabilistic Graphical Models • Reinforcement Learning • AI-Software

Abschluss: Computer Science, Master of Science
Stunden-/Tagessatz: anzeigen
Sprachkenntnisse: deutsch (Muttersprache) | englisch (verhandlungssicher)

Dateianlagen

resume_180324.pdf

Skills

My skill lies in planning, providing and executing full‐stack software projects, which solve real‐world problems, particularly those involving large amounts of data, statistical techniques and machine learning models. Initially specializing in reinforcement learning for robotics, I quickly ex‐ panded my expertise to encompass other time‐series‐related fields. These include supply chain management, financial trading systems, machine control systems, and satellite‐based monitoring solutions.

Projekthistorie

10/2023 - 01/2024
Inbound volume forecasting for warehouse (Software Lead / Machine Learning Scientist)
(>10.000 Mitarbeiter)

Implemented a forecasting model for inbound warehouse volume, in‐ tegrating exogenous and endogenous features, and connected it to SAP S/4HANA’s database. Developed a back‐test for this forecast using tech‐ nologies like Django, pandas, PostgreSQL, and Streamlit.

10/2021 - 10/2022
Reinforcement learning for auto focus (Machine Learning Engineer)
(Industrie und Maschinenbau, >10.000 Mitarbeiter)

Developed and fine‐tuned a reinforcement learning agent, focusing on feature engineering across dimensions like focus, contrast and brightness as well as stigmation. Designed an quality metric for rewards from images. Utilized technologies include Ray, RLLib, Pandas, Streamlit, Jupyter, and Gitlab CI.

02/2021 - 05/2022
Stock analysis platform (Software Lead / Machine Learning Scientist)
(Internet und Informationstechnologie, < 10 Mitarbeiter)

Developed and fine‐tuned a reinforcement learning agent, focusing on feature engineering across dimensions like focus, contrast, brightness, and stigmation, and designed an image quality metric for rewards. Uti‐ lized technologies including Ray, RLLib, Pandas, Streamlit, Jupyter, and Gitlab CI.

10/2020 - 09/2021
Earth observation platform (Machine Learning Engineer)
(Internet und Informationstechnologie, < 10 Mitarbeiter)

Developed a crop segmentation model to estimate crop availability in villages and created a download system for efficient access to Sentinel satellite’s geo‐spatial time‐series data. Advised on village detection us‐ ing a clustering method on pixel images, utilizing technologies like Post‐ GIS, FastAPI, Angular, Pandas, Numpy, scikit‐learn, and clearML.

12/2021 - 05/2021
Recommendation apps for books (Software Lead)
(Internet und Informationstechnologie, < 10 Mitarbeiter)

Implemented a baseline recommendation algorithm based on distance information retrieval and led the technical design for backend, frontend, and Android application. Utilized technologies include FastAPI, neo4j, PostgreSQL, Android, Flask, and scikit‐learn.

10/2018 - 04/2020
Alarm tracking system (Software Lead)
(Transport und Logistik, < 10 Mitarbeiter)


Received and refactored an old code base, focusing on the backend with PHP and mySQL, and optimized data transmission and energy us‐ age in an Android app’s tracking algorithm. Outsourced the Laravel‐ based frontend development and employed technologies like Android, PHP, Linux, Kalman Filter, Triangulation, and Microservices.

Reisebereitschaft

Weltweit verfügbar
Profilbild von Markus Semmler Senior • Deep Learning • Probabilistic Graphical Models • Reinforcement Learning • AI-Software aus Darmstadt Senior • Deep Learning • Probabilistic Graphical Models • Reinforcement Learning • AI-Software
Registrieren