Arthur Jochems

I am a PhD Data Scientist and AI Architect with 10+ years of experience deploying mission-critical modeling solutions across highly regulated sectors, including Life Sciences, Financial Technology, and Corporate Governance.

Through Modus Data

GmbH, I partner with lean technical teams, boutique asset managers, and innovative scale-ups to translate complex, unstructured data into strict operational advantages. My focus is bridging the gap between sophisticated deep learning architectures (LLMs, custom NLP pipelines, predictive models) and secure, production-ready MLOps.

Track Record of Value Delivered: Architected data pipelines driving €500k+ in revenue growth. Designed automated risk/compliance modeling resulting in €1M + in operational savings. Led end-to-end technical execution from experimental machine learning frameworks to secure cloud deployment (Azure/AWS/GCP). I operate primarily on a high-efficiency, fractional or outcome-based freelance model, providing strategic AI consulting and deep technical execution to Swiss and European enterprises.

Specialties: Large Language Models (LLMs), Natural Language Processing (NLP), MLOps, Data Science Strategy.

Principal AI Consultant & Founder at Modus Data GmbH (2026-06 – Present)

Modus Data GmbH is an independent Swiss data science and applied AI consultancy. We partner with lean technical teams, boutique asset managers, and regulated enterprises to bridge the gap between advanced deep learning architectures and production-grade MLOps pipelines.

As Principal

Consultant, I lead end-to-end strategy and technical execution for client initiatives, specializing in custom Large Language Model (LLM) implementation, advanced Natural Language Processing (NLP), and time-series forecasting frameworks.

• Available for fractional advisory roles, architectural audits, and end-to-end machine learning project delivery.
• Custom enterprise inquiries: info@modusdata.ch

Senior Data Science Consultant (via Modus Data GmbH) at Daiichi Sankyo Europe GmbH (2024-10 – Present)

• Training a foundational model on 100TB+ of digital pathology images
• Data analysis and model development for real-world evidence generation for various anti-cancer drugs.
• Analysis and model development based on medical imaging using computer vision/deep learning methods (ResNet, CovNeXT, ViT, vision transformer)
• Error detection in clinical study reports using LLMs (Qwen, huggingface transformers, langwatch, langchain, NLP).
• Development in Python, Scikit-Survival, Scikit-Learn, xgboost, survival modelling, Jupyter Lab, PyTorch, AWS EC2, AWS S3, FastAPI, streamlit, databricks, genie, instructor, pydantic, transformers, Docker, Google cloud run, Github

Senior AI Engineer (via Modus Data GmbH) at Mercedes-Benz Customer Assistance Center Maastricht NV (2026-04 – 2026-06)

Brought in as an external specialist for a targeted, 3-month project to architect and optimize production-ready AI workflow automation pipelines. Focused on advanced natural language processing (NLP) frameworks, including enterprise customer sentiment tracking, automated use-case summarization, and intelligent email classification.

• Engineered high-performance Databricks and Apache Spark data pipelines alongside deep query optimizations, increasing overall pipeline execution efficiency by 33%.
• Spearheaded the migration of the core codebase from native Azure infrastructure to Databricks, enabling a self-hosted LLM endpoint that slashed operational infrastructure costs by 50%.

Data Scientist/Generative AI Engineer at ZorgDomein (2024-06 – 2024-08)

• Development of NLP/generative AI algorithms to leverage large volumes of medical texts, using RAG, to create a recommender system for general practitioners looking for healthcare services.
• Development in Python, Azure Machine Learning, Azure Synapse Analytics, Azure Databricks, OpenAI, Langchain

Senior Data Scientist/Analyst at Bunge (2023-09 – 2024-01)

Development and maintenance of models and data pipelines related to marketing and sales of food products.

Senior Data Scientist at Roche (2022-11 – 2023-08)

• Development of NLP algorithms for the facilitation of tracking unplanned events and development and fine-tuning of Large Language Models (LLMs) for topic extraction of large pharmaceutical texts. These models reduce the time subject matter experts spend on drug production problem solving by over 70%.
• Development in Python, AWS Athena, Redshift, Sagemaker, SQL

Senior Data Scientist at T-Mobile Nederland (2021-05 – 2023-06)

• Developed machine learning models for predicting fraudulent orders. The B2C model has caught 2000 frauds over a period of 12 months, saving the company over €1M on fraudulent orders. The B2B model that I developed is expected to save the company over €3M yearly.
• Methods used: XGBoost, Random Forest, Logistic regression, Devops server, scikit-learn, kds. Development of ETL pipelines using Hive SQL, Hadoop, pyspark, great expectations data unit tests, pandas and Python. Dashboarding using PowerBI
• Agile scrum team environment. Direct communication with stakeholders.

Senior Data Scientist at PwC Deutschland (2020-08 – 2021-04)

• Developed an NLP machine learning solution that permits learning from financial documents without infringement on privacy. This solution is embedded in a SaaS solution called DataMesh. DataMesh is currently used within PwC to create their first generation of data products.
• Methods used: Synthetic data models using Generative Adversarial Networks (GANs). Development of Named Entity Recognition models using Deep Learning, BERT, Spark NLP. Development of langauge generation models using GPT2 and GPT3. Development of data models and ETL pipelines using SparkSQL, CamelK, Azure Machine Learning, Azure Data Factory, Azure Databricks, MS SQL. Dashboarding using PowerBI. Agile and Scrum.

Senior Data Scientist at Maastricht University (2014-09 – 2020-08)

Doctor of Philosophy (PhD) in Neuroscience – Ruhr-Universität Bochum (2010 – 2013)

Master of Science (MSc) in Computational Neuroscience – The University of Manchester (2009 – 2010)

Bachelor of Science (BSc) in Computer Science – Universiteit Utrecht (2007 – 2009)

– Marnix college Ede