Freelancer profile translated to English.
Back to original languageAbout Dimitrij
With over 12 years of experience, I develop AI and Machine Learning solutions for demanding applications in industry, automotive, MedTech, and life sciences. My focus is on areas where data is complex, systems are critical, and results must be practically applicable: Computer Vision, defect and anomaly detection, Edge AI, diagnostic ML methods, and predictive modeling.
I help companies implement AI initiatives with technical soundness and pragmatism – from feasibility assessment through prototypes and MVPs to integration into production systems. My experience ranges from live AI security systems and industrial quality inspection to patented AI/diagnostic methods based on high-dimensional blood measurement data in clinical study and validation phases.
WHAT I OFFER
- Feasibility checks and Proofs-of-Concept for Computer Vision, ML, and AI use cases
- Development of defect, anomaly, and object detection up to MVP and production deployment
- Building production-ready ML/Deep Learning pipelines with Python, PyTorch/TensorFlow, OpenCV, and MLflow
- Edge AI integration for camera systems, industrial plants, and embedded/NVIDIA Jetson setups
- AI system architecture: data strategy, model selection, validation, deployment, and scaling
- MedTech AI consulting: diagnostic AI, clinical data, SaMD-like product development, and technical documentation
- Technical leadership of AI projects, Fractional CTO, and sparring for product, management, and development teams
German
Native or bilingual
English
Fluent
Remote only
Primarily works remotely
Experience
- REHAU Industries SE & Co. KGData Scientist | Machine Learning EngineerMECHANICAL ENGINEERINGJanuary 2024 - December 2024 (11 months)Hamburg, GermanyAt Fujitsu, I was responsible as a Data Scientist for the development of Deep Learning models and the implementation of a Minimum Viable Product (MVP) as part of a partnership with REHAU Industries SE & Co. KG. The goal was to introduce an AI-based quality assurance system for polymer-based extrusion profiles, which are produced worldwide in high quantities and with great variety.I developed and trained Deep Learning models for automatic defect detection on production lines, validated them with real manufacturing data, and transferred the prototypes to an MVP. In the course of this project, I invented a method for AI-based anomaly detection, which was patented by Fujitsu under patent number WO2025180648A1.Press:
- EintrachtTech GmbHData Scientist | Machine Learning EngineerSPORTSJanuary 2023 - December 2023 (11 months)Frankfurt am Main, GermanyAt Fujitsu, I was involved as a Data Scientist in the development of an AI-based queue management system for the fan shop at Eintracht Frankfurt's Deutsche Bank Park. The project's goal was to analyze visitor flows at the fan shop based on data and predict the expected waiting time to enter the shop, in order to improve the fan experience and provide transparent waiting times.My focus was on the development of Computer Vision and Deep Learning models for analyzing camera data. For this, I developed tracking approaches based on Multiple Object Tracking, evaluated the positional changes of people in the camera image over time, and used the derived motion and pixel changes as the basis for predicting the queue time.Additionally, I developed LSTM-based time series models to derive a prediction of the expected waiting time from the movement patterns and changes in people's positions over multiple time steps. The models were subsequently integrated into an Edge AI environment on NVIDIA Jetson hardware by the project team and tested in a live pilot at the stadium surroundings.Press:
- RobotDreams GmbHSenior Data Scientist, Project & Team LeadBIOTECHSeptember 2022 - Today (3 years and 11 months)Graz, AustriaAs Co-Founder and technical project lead of RobotDreams GmbH, I was responsible for the technological development of a MedTech startup for diagnostic risk stratification in acute coronary syndrome. The technology is in the clinical study phase and is based on high-dimensional blood measurement data from hematology analyzers and ML models for ACS prediction.I built an international five-person core team and led the technical development. Together with cardiologists from the Medical University of Graz, I participated in the study protocol for the prospective study, which was approved by the Ethics Committee and registered on ClinicalTrials.gov under NCT06384846.My responsibilities include technical study management, coordination with study leadership and scientific staff, data preparation, statistical analysis, methods for automatic identification of blood cell populations in the high-dimensional event space, as well as training, validation, and testing of diagnostic AI models.Furthermore, I was significantly involved in raising approximately €1.1 million in R&D funding, including ESA BIC (€50,000), the Austrian Research Promotion Agency (FFG) (€224,000), and aws. I also invented and formulated several proprietary AI/diagnostic methods and worked with patent attorneys to draft their patent applications, including WO2023057334A1 and WO2024207041A1.Press:
Recommendations
Be the first to recommend Dimitrij
Help this freelancer shine by sharing your experience working together.
These freelancer profiles also match your criteria
Agatha Frydrych
Backend Java Software Engineer
4.7
(3)
2
Baptiste Duhen
Fullstack developer
4.6
(4)
5
Amed Hamou
Senior Lead Developer
4
(2)
7
Audrey Champion
Web developer
4.3
(3)
4
Education
- Bachelor of EngineeringHamburg University of Applied Sciences (HAW)2012Maschinenbau
- Master of ScienceHamburg University of Applied Sciences (HAW)2014Maschinenbau / Computational Engineering
Certifications
- Python ProgrammerDataCamp2019