Data scientist with a doctorate in quantitative ecology and ecological modelling and ten years of analysing multivariate time series from heterogeneous, incomplete sources. I build pipelines in which uncertainty and provenance are enforced: every retrieved payload is stored write-once with its hash, and every point estimate must lie inside its own Monte Carlo interval.
- grid-carbon-intensity: hourly carbon intensity of electricity in DE-LU and France, derived from ENTSO-E generation data with published life-cycle factors, and a 90% Monte Carlo interval at every step. DuckDB, dbt, GitHub Actions. In build; the capstone of my data-science training.
- eudr-geojson-validator: checks EUDR geolocation files against the European Commission's GeoJSON file description. v1.2
- heat-risk-hessen: heat, mortality and who lives where it's hottest in Hessen, 2000–2025. Hypotheses and data rules fixed before the data; built in February 2027
- soil-carbon-kriging: A Python pipeline that maps soil organic carbon and quantifies its uncertainty with ordinary kriging, using data from eight US Midwest fields and Germany’s BZE-LW soil inventory.
- eu-soil-crop-allocation: Which crop is grown where in EU agriculture, and what do lab-measured soil N, P, K and pH add beyond climate and terrain? A crop-allocation model on ~10,000 LUCAS 2018 points, tested against pre-registered hypotheses with spatial block cross-validation.
Python · R · SQL · dbt · DuckDB · Databricks . machine learning (scikit-learn) · agent-based modelling · Monte Carlo and global sensitivity analysis · geospatial data · Tableau . PowerBI
adewoleolagoke@gmail.com · LinkedIn · ORCID 0000-0002-0869-329X · Tableau Public
Open to data science and analytics roles in Germany.

