AI/ML Engineer focused on biomedical AI, protein modeling, and reproducible ML systems.

I build research-oriented machine learning pipelines for life science problems, with a focus on protein structure, medical imaging, and scalable experiment workflows.

• Biomedical AI and AI for science
• Protein binding-site prediction and structural bioinformatics
• Protein language models and transformer-based representation learning
• Reproducible ML systems with modern deep learning frameworks, configuration management, containerization, and experiment tracking
• Medical image representation learning and contrastive learning

Transformer fusion pipeline for protein binding-site prediction using ESM2 protein embeddings, physicochemical surface descriptors, and neighbourhood-aware attention.

• Developed as part of my Master's thesis in Artificial Intelligence at Johannes Kepler University Linz
• Combines protein language model embeddings, handcrafted surface descriptors, and local kNN attention
• Includes Hydra/Lightning training, Dockerized inference, benchmark evaluation, and production-style post-processing
• Provides model artifacts and an end-to-end CLI/Docker workflow for inference

Repository: https://github.com/lal3lu03/PockNet Model artifacts: https://huggingface.co/lal3lu03/PockNet

Medical vision-language learning project using contrastive learning for zero-shot chest X-ray classification.

• Bachelor thesis project
• Uses CLIP/CLOOB-style image-text representation learning
• Works with medical image preprocessing, report-derived labels, embeddings, and zero-shot evaluation

Repository: https://github.com/lal3lu03/MMCLFMI

MSc Artificial Intelligence, Johannes Kepler University Linz
Specialization: AI and Life Sciences

Before AI, I competed internationally as a professional judo athlete for Austria. That background shaped my discipline, resilience, and ability to execute under pressure.

Outside my main ML work, I maintain Android custom ROM builds for the OnePlus 8 Pro, which keeps me close to Linux-based systems, build tooling, and open-source device maintenance.

• Machine learning and deep learning
• Protein language models and transformer architectures
• Structural bioinformatics and drug-discovery-related modeling
• Medical imaging and multimodal representation learning
• Python, PyTorch, TensorFlow, Hydra, Docker, Linux, Git, W&B
• Multi-GPU training, reproducible experiments, and scientific evaluation workflows

LinkedIn: https://www.linkedin.com/in/maximilian-hageneder-ai