Senior Data Scientist, Probabilistic Modeling – Architecture, Adjacent

This is a fully remote position, open to applicants in Romania.

📋 Description

• Create and assess probabilistic models for dynamic pricing, shipping cost assessment, recommendations, and segmentation.

• Design and execute Bayesian statistical models, encompassing priors, likelihoods, and posterior inference.

• Develop formulations for Markov chains and Hidden Markov Models to analyze sequential and behavioral patterns.

• Utilize MCMC techniques, such as Metropolis-Hastings sampling, and verify convergence and sampling quality.

• Construct mixture models, particularly Gaussian Mixture Models, for customer or product segmentation purposes.

• Employ Expectation-Maximization for estimating latent variables and conducting unsupervised learning tasks.

• Work closely with the solution architect and the client's CTO to ensure model design aligns with platform architecture.

• Provide guidance to backend engineering on production service architecture, API design, data contracts, and microservices/event-driven integration.

• Define methodologies for model training, validation, versioning, monitoring, drift detection, and retraining.

• Collaborate with delivery and engineering leaders to size, sequence, and estimate modeling projects.

• Document modeling assumptions, methodologies, and validation outcomes.

• Offer handoff support for the ongoing maintenance by the engineering team.

⛳️ Requirements

• Proficient in English with 90% written and oral skills, at least at a B2 level.

• Significant experience in effectively communicating with both technical and business stakeholders, including discussions at the CTO level.

• Proven expertise in Bayesian statistical models, Markov chains, Hidden Markov Models, MCMC methods including Metropolis-Hastings sampling, mixture models (preferably Gaussian Mixture Models), and Expectation-Maximization.

• Strong preference for experience in designing statistical/ML models with an eye towards production deployment; hands-on production implementation is advantageous but not essential.

• Proficient in Python or R.

• Familiarity with probabilistic/statistical libraries such as PyMC, Stan, scikit-learn, NumPy, or SciPy.

• Capability to convert statistical/mathematical models into service-oriented production architecture.

• Understanding of APIs, data contracts, collaboration with backend engineering, and architectural integration.

• Solid grasp of version control, testing practices, and CI/CD processes.

• Strong written and verbal communication abilities.

• Preferred experience in e-commerce or retail, particularly in areas like pricing optimization, customer segmentation, or demand forecasting.

• Familiarity with REST/GraphQL microservices and event-driven systems is a plus.

• Knowledge of .NET, Java, or Node.js backend ecosystems is advantageous.

• Exposure to MLOps concepts such as model registries, monitoring, or feature stores is favored.

• Preferred background in pricing science, recommendation systems, or marketing analytics.

• Experience in conveying modeling recommendations to business or executive stakeholders is preferred.

🏝️ Benefits

• Option for remote work.

• Exceptional work environment recognized by Great Place To Work.

• Diverse company culture.

• Opportunity to engage in a client project lasting 3–6 months.