Fatima Zahra
Data Scientist
Amsterdam, Netherlands·fatima.zahra@example.com·+31 20 555 0174·linkedin.com/in/fatimazahra·github.com/fatimazahra
Summary
Data scientist with 5 years building forecasting, experimentation and recommendation models for marketplaces. Takes models from exploration to production, and judges success by business metrics rather than offline accuracy alone. Enjoys partnering with product and operations teams, explaining models in plain language and teaching good experimentation.
Experience
Data Scientist | Tulip Market | Amsterdam | Mar 2021 – Present
- Built a demand-forecasting model in PySpark that cut stockouts by 22% across 8,000 products.
- Designed an experimentation platform that ran 150 A/B tests a year with consistent guardrail metrics.
- Developed a recommendation model that increased basket size by 6%, worth €3M a year.
- Deployed 5 models to production with MLflow and monitoring, with zero rollbacks in 2024.
- Mentored 3 analysts moving into data science through weekly pairing sessions.
- Built a causal-inference toolkit for marketing measurement adopted by 4 analytics teams.
Data Analyst | Route24 | Rotterdam | Jan 2019 – Feb 2021
- Analyzed 4M deliveries a month to optimize courier routing, saving €600K a year.
- Built churn prediction for B2B customers that helped account managers retain €1.2M in revenue.
- Created self-serve dashboards in Looker used weekly by 120 colleagues.
- Automated 10 recurring reports with SQL and Airflow, saving 15 hours a week.
- Designed a pricing experiment for delivery fees that increased margin per order by 4% with no loss in volume.
- Presented quarterly model reviews to the product leadership team, prioritizing 6 new data products.
- Built anomaly detection for order data that flagged 3 major incidents within minutes.
Research Assistant | University of Amsterdam | Amsterdam | Sep 2017 – Dec 2018
- Analyzed survey data from 12,000 respondents for a published labour-economics study.
- Wrote reproducible R pipelines adopted by 4 researchers in the department.
- Presented results at 2 academic workshops on applied econometrics.
Projects
Open forecasting benchmarks · github.com/fatimazahra/bench | Contributor | Jan 2022
- Contributed demand-forecasting benchmarks to an open-source library used by 2,000+ data scientists.
Education
University of Amsterdam | M.Sc., Data Science | 2016 – 2018
Skills
Data science: Forecasting, Experimentation, Recommendation systems, Causal inference
Tools: Python, PySpark, SQL, MLflow, Airflow, Databricks
Languages
English: Fluent (C2) · Arabic (Native) · Dutch: Professional (B2)