A data scientist who only lists Kaggle competitions and notebook experiments looks like a hobbyist. A data scientist who shows models in production with measurable business impact looks like a hire. The distinction matters more in 2026 than it did three years ago.
Top: domain, production experience, stakeholders
Two-sentence profile summary positioning your domain (fintech, healthcare, marketing, retail), whether you have models in production, and your stakeholder mix (business teams, product, leadership).
Experience with deployed work
Per role: company, team type (embedded in business unit vs. central DS team), problems tackled, models built, and crucially — did they go to production. "Built churn model" is weak. "Built churn model deployed via Vertex AI, serving 12M predictions/month, drove €600k retention budget reallocation Q3 2024" is strong.
Stack: where you are on the production spectrum
For research-leaning: Python, scikit-learn, statsmodels, jupyter, R. For ML-engineering-leaning: PyTorch/TensorFlow, MLflow, Kubeflow, FastAPI, Docker, Kubernetes. For productionised work: feature stores (Feast, Tecton), monitoring (Evidently, Arize). Be honest about where you live.
Concrete example bullets
- Built churn prediction model (XGBoost + logistic baseline) deployed via Vertex AI; 12M predictions/month; €600k retention budget reallocation Q3 2024.
- Led recommender system rebuild for marketplace; A/B-tested two variants; winner uplifted GMV per session by 8.4%.
- Owned NLP-based ticket-routing classifier; reduced first-response time by 34%.
- Stack: Python (pandas, scikit-learn, PyTorch), GCP Vertex AI, BigQuery, Feast, MLflow, Docker.
- MSc Statistics (Leiden, 2019); Deeplearning.AI specialization (2022).
Pitfalls for this role
- Kaggle competitions as main proof. Useful but secondary to deployed work.
- Algorithm names without business outcome. "Used XGBoost" — for what, with what result?
- Production claims you can't defend. Hiring data leads will probe quickly.
- No deployment context. A model in a notebook is half the job.
Substance over style. A calm layout lets the work speak.
Salary indication 2026 (NL): junior DS €45–60k, mid €60–85k, senior €85–115k, staff/principal €115k+.
Read next
- Tailoring a CV in two minutes — how it works without the cringeThe "one CV for everything" advice is dead. Here's how to make a job-specific CV in two minutes without fighting Word.
- ATS systems in 2026 — what actually works, what's mythNo PDF panic, no table phobia, no nonsense about white space. A sober look at how modern ATS systems process your CV in 2026 — and what really matters.