Skip to content

Dr. Martijn Arts

Freelance Data Engineer · Python Software Engineer · Data Scientist

Freelancer for Data Engineering, Python Software Engineering and Data Science, backed by a strong machine learning and statistics background. Industry experience in consulting and in leading data projects through the entire product life cycle — from ideation and proof-of-concept to bringing data products into production in the cloud and operating them.

Core Skills

What I work with

Languages & core
Python (Pandas, Polars, NumPy, Pydantic, FastAPI), SQL, Bash
Data & cloud
Snowflake, dbt, Postgres, Spark / Databricks, AWS (S3, Batch, Lambda, Step Functions, EventBridge Scheduler, CloudFormation), Azure (AzureML, DevOps Pipelines, Application Insights), Docker, Kubernetes, Datadog
Machine learning
Scikit-learn, XGBoost, Prophet, BoTorch, Scikit-Optimize, MLflow — statistical modelling, time-series forecasting, black-box optimization, model calibration & evaluation
AI-assisted development
GitHub Copilot, Claude Code, Anthropic (Claude) and OpenAI models — code migration, refactoring, test generation and day-to-day development
Engineering practice
Unit & integration testing (Pytest), CI / CD, Git, MLOps, monitoring & alerting, code review, refactoring and stabilization of large code bases
Ways of working
Agile team lead (SCRUM), requirement & stakeholder management, project management, mentoring, presenting to executive and non-technical audiences

Selected Work

How I help

From prototype to production

I take data products past the proof-of-concept stage and keep them running. A predictive credit risk assessment model started as a PoC and went on to be operated as a fully productionized data product covering a receivables portfolio of several million items. More recently I designed and built the backend of an energy market report on Python, Snowflake and AWS, which replaced several days of manual work per update and enabled dynamic report updates.

Take-over, clean-up and stabilization

Inherited code bases are a large part of what I do. On an orchestration system for an energy market simulation model, I took the project over and reduced production incidents from daily to one or two per week within less than a month — by increasing system transparency through logging improvements and reduction of technical debt, rather than by rewriting it.

Migration to modern Python

I move projects off legacy stacks without losing behaviour along the way: a large, monolithic C++ component migrated to Python with the help of LLM-assisted development, and more than 50 data and business processes migrated from a legacy system onto a new data platform, working in a Python code base of over 100,000 lines across several agile teams.

Let’s make your data project succeed.

Whether you need help with creating a new data product, taking the next step with an existing one or migrating a legacy stack into modern Python — I am available for freelance projects, remote with occasional on-site visits within Germany. Tell me about your project and I will tell you honestly whether I am the right fit.