Data Scientist
Designs the statistical and machine learning models that turn your data into predictions and measurable optimizations.
Why this hire matters.
The data scientist creates value where business rules hit their ceiling: scoring, demand forecasting, fraud detection, pricing, recommendation. Hired well, they attack high-ROI problems with the right method — sometimes a simple regression beats an expensive deep learning model. Hired poorly, you get the well-known syndrome: brilliant notebooks that never leave the laptop, POCs stacking up without ever touching production or the P&L. The decisive criterion is no longer model sophistication but product sense: understanding the business problem, choosing the simplest solution that works, and collaborating with engineers to ship it. In the LLM era, their ground shifts to evaluation, proprietary data and specialized models — where the generic API isn't enough.
Key missions.
- Frame business problems as modeling problems
- Explore data and build features
- Develop and evaluate models (classical ML, deep learning)
- Measure real impact (A/B tests, business metrics)
- Partner with ML engineers on industrialization
- Communicate results and limits to decision-makers
- Track methods (LLMs, causal inference, etc.)
Skills.
Technical skills
- Python (pandas, scikit-learn, PyTorch)
- Statistics and experimentation
- SQL
- Model evaluation
- Feature engineering
Expected qualities
- Product sense
- Scientific mindset
- Communication
- Pragmatism
Common stack
Salaries 2025-2026
| Level | Experience | Annual gross base |
|---|---|---|
| Junior | 0-2 yrs | 44–52 k€ |
| Mid-level | 2-5 yrs | 52–70 k€ |
| Senior | 5-8 yrs | 70–85 k€ |
| Lead / Staff | 8+ yrs | 85–130 k€ |
Paris market ranges, 2025-2026.
Outside the Paris region, expect 10 to 20 % less.
Sources : Factoriel 2026 · Silkhom 2026 · liora 2026 · Michael Page 2026
Hiring this profile.
Typical background
Master's/PhD in applied math, stats or CS, or engineering school. Models shipped to production weigh more than publications.
When to hire
When a high-stakes problem resists business rules and you have the data (and a data engineer) to attack it — often Series A/B.
Career path
Senior → Lead Data Science → Head of Data Science / Chief Data & AI Officer, or a move to ML engineer / AI researcher.