07/20/2026 updated
100 % available
AI / ML & Data Engineer
copenhagen, Denmark
Worldwide
ML Specialization – Stanford / DeepLearning.AI, 2023About me
AI/ML & Data Engineer with hands-on experience building production SaaS platforms. Skilled in Python, TensorFlow, GCP, and AWS. Built EnergyLens (Nordic energy forecasting) and MarketLensPro (trading analytics). Stanford ML certified. Based in Copenhagen, available for remote and on-site projects.
MLflowMLOpsGCPData AnalysisArtificial Neural NetworksGitHubPython (Programming Language)Docker ContainerFeature EngineeringPySparkGoogle BigqueryApache Beam
Machine Learning & Deep Learning
Expertise in building production ML systems using PyTorch, TensorFlow, XGBoost, Scikit-learn, SHAP, and Optuna. Development of 8-model neural ensembles combining deep-learning and gradient-boosting models for real-time signal generation across multiple asset classes.
Time Series Forecasting
Application of advanced time series architectures including Transformer, CNN-LSTM, TCN, N-BEATS, and LSTM-GRU ensembles for tasks such as day-ahead electricity price forecasting with measured accuracy via MAE, RMSE, and directional accuracy metrics.
Data Engineering on GCP
Hands-on data engineering using BigQuery, PostgreSQL, dbt, PySpark, Airflow, Pub/Sub, Firestore, and DuckDB on Google Cloud Platform. Design and operation of bitemporal data pipelines and fully automated MLOps workflows including drift detection and model hot-reload.
Cloud & DevOps
Deployment and orchestration of production services using GCP (Cloud Run, GCS, Scheduler), Docker, CI/CD pipelines, and GitHub Actions for zero-downtime model versioning and automated retraining.
FastAPI & Backend Development
Development of B2B APIs and backend services using FastAPI and Python, including HMAC-SHA256 authentication, Firestore rate limiting, and live signal delivery for SaaS platforms.
RAG & Vector Search
Implementation of Retrieval-Augmented Generation chatbots using Gemini embeddings (768-dim), Firestore cosine KNN vector search, and LLM generation over multi-document knowledge bases.
Product & Full-Stack Integration
End-to-end product ownership including React frontends, Stripe billing, Firebase Auth, Telegram alerts, paper trading engines, and cTrader Store plugin integration.
AutoML & Anomaly Detection
Building AutoML education platforms with 9+ algorithms, live market data, and walk-forward backtesting using SHAP, as well as weighted ensemble anomaly detectors combining statistical, ML, and autoencoder methods with drift monitoring and regime detection.
SQL & Python
Proficient use of Python (~19,000 lines in production) and SQL for data manipulation, pipeline development, model training, and deployment across multiple live platforms.
Expertise in building production ML systems using PyTorch, TensorFlow, XGBoost, Scikit-learn, SHAP, and Optuna. Development of 8-model neural ensembles combining deep-learning and gradient-boosting models for real-time signal generation across multiple asset classes.
Time Series Forecasting
Application of advanced time series architectures including Transformer, CNN-LSTM, TCN, N-BEATS, and LSTM-GRU ensembles for tasks such as day-ahead electricity price forecasting with measured accuracy via MAE, RMSE, and directional accuracy metrics.
Data Engineering on GCP
Hands-on data engineering using BigQuery, PostgreSQL, dbt, PySpark, Airflow, Pub/Sub, Firestore, and DuckDB on Google Cloud Platform. Design and operation of bitemporal data pipelines and fully automated MLOps workflows including drift detection and model hot-reload.
Cloud & DevOps
Deployment and orchestration of production services using GCP (Cloud Run, GCS, Scheduler), Docker, CI/CD pipelines, and GitHub Actions for zero-downtime model versioning and automated retraining.
FastAPI & Backend Development
Development of B2B APIs and backend services using FastAPI and Python, including HMAC-SHA256 authentication, Firestore rate limiting, and live signal delivery for SaaS platforms.
RAG & Vector Search
Implementation of Retrieval-Augmented Generation chatbots using Gemini embeddings (768-dim), Firestore cosine KNN vector search, and LLM generation over multi-document knowledge bases.
Product & Full-Stack Integration
End-to-end product ownership including React frontends, Stripe billing, Firebase Auth, Telegram alerts, paper trading engines, and cTrader Store plugin integration.
AutoML & Anomaly Detection
Building AutoML education platforms with 9+ algorithms, live market data, and walk-forward backtesting using SHAP, as well as weighted ensemble anomaly detectors combining statistical, ML, and autoencoder methods with drift monitoring and regime detection.
SQL & Python
Proficient use of Python (~19,000 lines in production) and SQL for data manipulation, pipeline development, model training, and deployment across multiple live platforms.
Experience
Professional experience0 - 5 years
Freelancer since2023
Languages
DanishFluentEnglishFluent
Project history
Full end-to-end ownership of a live SaaS trading-analytics platform (~19,000 lines Python) on GCP Cloud Run. Built an 8-model ensemble generating real-time ML signals across 19 asset classes, automated MLOps with drift detection and GPU retraining, a RAG chatbot using Gemini embeddings and Firestore vector search, a B2B API with HMAC-SHA256 auth, and full-stack integration including Firebase Auth, Stripe billing, Telegram alerts, and a cTrader Store plugin.
Development and operation of a live Nordic electricity-price forecasting platform on GCP Cloud Run using Python/FastAPI and React. Built an 8-model neural ensemble predicting DK1/DK2 day-ahead prices 24h ahead over Nord Pool, ENTSO-E, and weather data. Implemented automated scoring, 5-gate signal-quality validation with SHAP explainability, and a leakage-free bitemporal backtesting data model.
Developed a weighted ensemble of statistical, ML, and autoencoder detectors with drift monitoring and regime detection capabilities.