Tech & product example
Data science resume example
A data science resume example for ML and analytics roles in Zürich and continental Europe. Four years of production models, experimentation and a publications block — the impact and research language DS hiring managers expect before case studies.
Pourquoi ce CV est structuré ainsi
Yuki has four years post-MSc at ETH — experience leads with model impact metrics (fill rate, conversion lift) and production pipelines, not notebook screenshots. Publications live in a custom section so peer-reviewed work scans without crowding role bullets. ETH MSc and thesis sit in education; PhD is not claimed. Minimal template keeps dense metrics readable and ATS-safe. Skills split methods (forecasting, experimentation) from tools (Python, dbt). No photo. Research and industry bullets balance for roles spanning applied ML and analytics.
Professional Summary
Data scientist with four years building forecasting and experimentation systems for marketplace products. Combines production ML pipelines with peer-reviewed research in demand modelling.
Experience
- Built demand forecasts that improved inventory fill rate by 6pp across 12 SKU categories
- Designed A/B tests for pricing; shipped winner lifting conversion 4.2% with €1.1m annualised impact
- Productionised feature pipelines in Python/SQL used by three product squads
- Reduced model retraining cycle from weekly to daily via Airflow orchestration
- Built churn propensity model improving retention campaign targeting precision 22%
- Delivered Looker dashboards adopted by commercial team for weekly planning
Publications
- Co-author; introduced embedding-based cold-start method for new SKUs
- Presented synthetic control methodology adopted in production A/B framework
Education
Grade 5.6/6 · Machine Learning, Causal Inference, Optimisation
Skills
Languages
Minimal template · see layout
Ce que regardent les recruteurs
- Model impact in business terms — conversion, fill rate, revenue
- Experimentation design and shipped A/B test outcomes
- Production pipelines, not only Jupyter notebooks
- Publications or talks when applying to research-heavy teams
- ETH or strong MSc credentials in education, not headline
- Tools tied to workflows (dbt, Airflow) not buzzword lists
- One-page default for four years applied DS
Erreurs fréquentes
- Listing sklearn tutorials without production deployment
- Burying publications when applying to research-adjacent roles
- Vague “built models” without business metrics
- 30-algorithm skills dump with no pipeline ownership
- Leading with education when four years of impact exist
Mots-clés ATS
| Keyword | Why it matters |
|---|---|
| Experimentation | A/B testing |
| Forecasting | Demand modelling |
| Python | Core language |
| SQL | Data access |
| Production ML | Deployment signal |
| Causal inference | Methods depth |
| dbt | Analytics engineering |
| Publications | Research signal |
Questions fréquentes
Yes for research-heavy or applied ML roles. A compact publications block beats long paper abstracts in experience.