EN DE

Capital Markets Analytics & Machine Learning

Two decades in capital markets, now paired with machine learning and Python: I build new models on new technology and mature them into commercial products — together with domain experts who know the markets from the inside.

Get in Touch GitHub

From Finance
to Data Science

Two decades in the financial sector taught me how markets actually behave — and where data makes the difference in strategic decisions. Today I put that experience to work in model development, using machine learning and Python.

My focus is on quantitative capital-market analytics, time-series & signal detection, and full-stack web applications: new models built on emerging technology, designed from the start to answer a real question rather than to look good in a backtest.

I rarely build alone. Working closely with domain experts — in mining, digital assets, or capital markets — is what turns a promising prototype into something that holds up in practice and matures all the way to commercialization.

Based between the coasts of Schleswig-Holstein and Berlin's tech ecosystem, I build production-grade data systems, from ETL pipelines to interactive dashboards, with an emphasis on explainability, robustness, and business impact.

Recent Work

📉 Daily Dip
Radar

Daily Dip Radar – Pipeline

End-to-end capital-market screening system that scans a stock universe for dip and trend setups. Automated EOD data pipeline (PostgreSQL + scheduler), technical-signal engine, and an event-driven dashboard that refreshes after each run.

Python PostgreSQL Pandas Technical Analysis
View Project →

📊 HYPE Crypto
Treasury

HYPE Crypto Treasury – Engine

Valuation and risk engine for single-asset digital-asset treasury vehicles, initially focused on the Hyperliquid ecosystem. Combines on-chain data, market data, and SEC filings into fundamentals, liquidity, risk, and mNAV modules — extended by scenario and Monte Carlo stress testing.

Python NumPy statsmodels Risk Analytics
View Project →

⛏️ exploreIQ
Platform

exploreIQ – Pipeline

European exploration intelligence platform: open geodata from national geological surveys, mining-claim registries, and Sentinel-2 satellite imagery unified into a single spatial model. Historical exploration reports are mined via OCR, and an Exploration Opportunity Score (0–100) rates every 1 km² grid cell on an interactive map.

GeoPandas PostGIS Streamlit Remote Sensing
View Project →

Domains & Tech Stack

📈 Quantitative Market Analytics

  • Time-Series & Signal Detection
  • Dip & Trend Screening
  • Backtesting & Evaluation
Pandas NumPy EMA / RSI Backtesting

🤖 Advanced Machine Learning

  • Feature Engineering & Selection
  • Ensemble Methods & Cross-Validation
  • Model Explainability
Scikit-learn XGBoost SHAP statsmodels

💼 Financial Analytics

  • Portfolio Optimization
  • Risk Modeling & Mitigation
  • Predictive Analytics & BI
Python SciPy Power BI SQL

🌐 Full-Stack Web Applications

  • Interactive Trading Dashboards
  • Event-Driven Data Views
  • Real-Time Visualization
Reflex Streamlit FastAPI Plotly

📊 Data Engineering & ETL

  • ETL & Market-Data Pipelines
  • Scheduling & Automation
  • Scalable Data Architecture
PostgreSQL MongoDB SQL Git / GitHub

🗺️ Geospatial Data Science

  • Spatial Analysis & Clustering
  • GIS Data Processing
  • Interactive Mapping
GeoPandas PostGIS MapLibre Rasterio
Frank Hasdorf

Freelance Financial Data Scientist

Berlin & Schleswig-Holstein, Germany

Let's Build Something
Meaningful Together.

Whether you need quantitative market analytics, advanced ML models, or a full-stack data application — let's explore what's possible.

Send Message
GitHub Repositories