AI Research Engineer / ML Researcher.

Conceptualization, implementation, and scientific validation of AI core innovations.

Development of novel generative image models and training pipelines for architecture-specific application scenarios.

Research on controllable, region-, and material-based generation methods as a unique selling point of the platform.

Conducting systematic experiments, benchmarking, and qualitative and quantitative evaluations.

Closely working with product and engineering teams to transfer research results into production-ready AI systems.

Responsibility for the scientific substance of research projects (e.g., within EXIST, ZIM, or other funding programs).

Positioning as the scientific core of AI research within the team.

Aufgaben

Conceptualization and further development of the KI-Kernarchitektur for generative image models with a focus on architecture-specific application scenarios.

Development, training, and optimization of generative deep-learning models (especially Diffusion Models, UNet architectures, Transformer-based approaches, Latent Diffusion).

Implementation and evaluation of Fine-Tuning- und Adaptionsmethoden like LoRA, DreamBooth-like procedures, and domain-specific transfer learning.

Research on controllable generation mechanisms, e.g., region-based control, material-, and text-based conditioning, as well as structure- and context-aware image synthesis.

Building and maintaining scalable trainings- und experimentierpipelines (e.g., PyTorch, GPU clusters, automated experiments).

Systematic wissenschaftliche Validierung of the developed models through benchmarking, ablation studies, and qualitative and quantitative evaluation methods.

Analysis, preparation, and curation of trainings- und evaluierungsdatensätzen in an architectural context.

Dokumentation of research results and methodological decisions to ensure reproducibility and knowledge transfer.

Closely working with the product-, frontend-, and backend-team, transferring research results into production-ready AI systems.

Technical support in applying for and conducting research and funding projects (e.g., EXIST, ZIM, EU programs).

Observation and evaluation of current Forschungsentwicklungen im Bereich generativer KI and deriving relevant innovation potential.

Qualifikation

A completed Master- or PhD-study in Informatik, Mathematik, Data Science, Computational Design, or a comparable field.

• Zwingende Expertise in generativer KI für Bilder, especially in training deep-learning models.
• Nachweisbare praktische und/oder wissenschaftliche Erfahrung with Diffusion Models, Transformers, UNet architectures, Latent Diffusion Models, and LoRA.
• A very good understanding of Deep-Learning-Methods, model architectures, training strategies, and evaluation methods.
• Experience in handling Forschungs- und Experimentier-Workflows, reproducibility, and systematic model evaluation.
• A strong analytical thinking ability and interest in the connection between wissenschaftliche Forschung und produktnaher Anwendung.
• Self-sufficient, structured work style and high intrinsic motivation to explore new AI approaches and bring them to maturity.

Benefits

A secured and long-term employment perspective through funding.

- Direct influence on a future-oriented deep-tech research project.

- Equity: The allocation of company shares is negotiable and part of the compensation package. Chance for CTO role.

- Competitive remuneration (negotiable).

- Office space in Cologne

Start date is expected to be September 2026

Application deadline is January 10, 2026

Founder: Han Cheol Yi https://www.linkedin.com/in/h-yi/

Find Jobs in Germany on Arbeitnow