Data Scientist – MMO Enhancements

Satori Analytics

Data Scientist at Satori Analytics focusing on AI-driven analytics enhancements across industries. Responsible for Bayesian modeling, causal inference, and collaborative problem-solving in data science.

Posted 7/10/2026full-timeAthens • 🇬🇷 GreeceMid-LevelSeniorWebsite

Core Competencies

Role fit

Core Competencies

Use this summary to align your resume positioning with the role.

Demonstrates expertise in Bayesian modeling, including the ability to specify models from scratch and handle uncertainty propagation. Proficient in Python and SQL, with a strong foundation in applied statistics and causal inference methodologies.

Highest-signal resume keywords

Bayesian ModellingPython ProgrammingApplied StatisticsCausal InferenceSQL Proficiency

ATS Keywords

Tailor your resume

Applicant Tracking System Keywords

Tip: use these terms in your resume and cover letter to boost ATS matches.

Hard Skills

Bayesian RegressionMCMC SamplingHierarchical StructuresModel EvaluationOptimisation TechniquesCausal Ground TruthChange-Point DetectionStatistical DistributionsHypothesis TestingModel Diagnostics

Soft Skills

Clear Written CommunicationIndependenceProactive Problem Solving

Tools & Technologies

PyMCStanScipy.optimizeCVXPYGitArviZCausalImpactMLflowNumPyroPyro

Industry Keywords

Causal InferenceGeo ExperimentsMedia DynamicsAdstock TransformationsMulti-Objective Optimisation

Tech Stack

Tools & technologies

NumpyPandasPythonScikit-LearnSQL

About the role

Key responsibilities & impact

**What Your Day Might Look Like:**
- Specify Bayesian regression models from scratch — likelihood, priors, hierarchical structure — not just library defaults
- Build and diagnose models in PyMC and/or Stan — MCMC sampling, convergence diagnostics (R-hat, ESS, divergences), and posterior visualisation with ArviZ
- Design hierarchical structures with partial pooling and multi-level architectures, handling sparse group data gracefully
- Chain models correctly — sequential, multi-stage architectures with full uncertainty propagation (Monte Carlo through the chain, never point estimates)
- Model media dynamics — adstock and saturation transformations (geometric, Weibull, Hill), with parameters specified via priors and response curves extracted from posteriors
- Optimise budgets under constraints — scipy.optimize and CVXPY for channel floors/ceilings and portfolio constraints, plus multi-objective work (Pareto frontiers, weighted utility, conflicting objectives)
- Establish causal ground truth — geo experiment design and analysis, Difference-in-Differences, Synthetic Control, and experiment-to-model calibration
- Handle structural breaks — piecewise regression, change-point detection (PELT, BOCPD, Bayesian), and regression discontinuity design
- Document your work — methodology docs, assumption logs, and structured write-ups that stand up to scrutiny

Requirements

What you’ll need

**Your Superpowers🚀:**
- Strong Python — production-quality, clean, reproducible code (pandas, NumPy, scikit-learn)
- Solid SQL — joins, window functions, and aggregations at scale
- Applied statistics depth — distributions, regression, hypothesis testing, model evaluation, and genuine comfort working with probability distributions and uncertainty
- From-scratch Bayesian modelling — you can write down a likelihood, choose priors, and defend a hierarchy, not just call a packaged tool
- PyMC and/or Stan fluency — including reading and acting on convergence diagnostics
- Uncertainty discipline — you propagate posteriors through model chains rather than collapsing to point estimates
- Optimisation experience — constrained nonlinear and multi-objective problems in a real-world setting
- Causal inference toolkit — geo experiments, DiD, Synthetic Control, and calibrating models against experimental results
- Git / version control as a matter of habit
- Clear written communication — you document methodology and assumptions without being asked
- Independence — you work from a defined brief, manage your own progress, and are proactive when blocked
- AI awareness — you follow the space and know when (and when not) to reach for AI tooling
**Bonus Points for:**
- CausalImpact or equivalent Bayesian causal inference frameworks
- MLflow or other experiment-tracking tooling
- NumPyro / Pyro experience
- Geo lift tests — experience running or analysing them in a media context

Benefits

Comp & perks

**Perks on Perks:**
- Competitive salary and hybrid work model – come hang out in our Athens office or work remotely from anywhere in European economic Area (EU, Switzerland etc.) or UK (up to 6 weeks per year).
- Training budget to level up your skills from the top tech partners in the market (Microsoft, AWS, Salesforce, Databricks etc.) – whether it’s certifications or courses, we’ve got you covered.
- Private insurance, top-tier tech gear, and the chance to work with a stellar crew.