Candidate
Jobs / Data in Johannesburg / Samaha Consulting

Data Scientist

Posted 1 week ago

Samaha Consulting · Johannesburg
Apply on the original posting ↗You'll apply on the original posting. We keep this job in your tracker so you never lose the thread.

Key Responsibilities

Data Modelling & ML Development

  • Build and enhance personalised recommendation models for the personalised inbound product platform.
  • Develop models for customer propensity, product selection, ranking, pricing and affordability optimisation.
  • Evolve existing scoring approaches towards adaptive, data-driven ML and AI decisioning.
  • Develop customer-level features using usage, spend, recharge, purchase, engagement and response behaviour.
  • Develop optimisation approaches that balance customer relevance, affordability, conversion and commercial value.
  • Build, train, validate and optimise predictive and recommendation models using appropriate statistical and machine-learning techniques.
  • Develop learning-to-rank, experimentation, uplift and adaptive-learning approaches where appropriate.
  • Design robust control groups, champion/challenger tests and experimentation frameworks.
  • Establish appropriate measures of model performance, including predictive accuracy, lift, calibration and incremental commercial impact.
  • Build model explainability and interpretability into recommendation decisioning.
  • Monitor model performance, feature drift and model degradation over time.
  • Define modelling standards, review technical approaches and provide design assurance for work delivered by other data scientists.
  • Select fit-for-purpose algorithms and establish robust champion/challenger, back-testing and model validation approaches.

Data Analysis & Insight Generation

  • Perform advanced analysis (e.g., cluster profiling, cohort analysis, penetration/lift analysis) at scale using Spark/Databricks on large datasets (10M+ rows).
  • Generate business-friendly insights and recommendations from model outputs for stakeholders.
  • Query and manipulate data using SQL (Hive/Trino) and Python (pandas, PySpark).
  • Lead complex analytical investigations across multiple data domains and convert findings into clear commercial decisions and measurable actions.
  • Establish reusable analytical frameworks, curated features and quality checks that improve consistency and speed across the team

MLOps & Productionisation

  • Own and lead the transition of models from development into production, including monitoring, regression testing and performance evidence.
  • Apply and enforce MLOps best practices (version control, experiment tracking, CI/CD, model monitoring) in partnership with the AI Ops function.
  • Track and report on the performance and drift of models deployed in production.
  • MLOps capability is mandatory: independently implement reproducible training pipelines, experiment tracking, model registration, automated testing and controlled deployment workflows.
  • Define production-readiness criteria covering data quality, model validation, explainability, security, documentation, rollback and operational support.
  • Establish monitoring for data drift, concept drift, model performance, service reliability and business outcomes, with clear alerting and remediation processes.
  • Drive model lifecycle governance, technical documentation and audit evidence in partnership with AI Ops, data engineering, architecture and risk stakeholders.

Collaboration & Capability Building

  • Support upskilling initiatives (MLOps, coding standards, Agile/Scrum for data science) and mentor graduates/interns.
  • Provide technical leadership, coaching and structured peer review to data scientists, graduates and cross-functional delivery teams.
  • Lead solution design across data science, data engineering, MLOps, BI and architecture, resolving dependencies and technical risks early.
  • Improve team capability through

Job Specifications — Minimum & Preferred Requirements

Education

  • Minimum 3-year tertiary degree in a STEM field: Computer Science, Engineering, Mathematics, Statistics, Data Science or a related quantitative discipline.
  • Honours or Master's degree in Data Science, Statistics or a related field advantageous.
  • Fluent in English.

Experience

  • 5-10 years' experience in data science / advanced analytics, including ownership of complex solutions from business problem framing through production deployment and value realisation.
  • Experience working with large-scale distributed data processing (Spark/Databricks) is highly advantageous.
  • Demonstrated hands-on MLOps experience and a track record of deploying and operating machine learning models in production are mandatory.
  • Experience with model governance, explainability, risk controls, data quality and production monitoring in an enterprise environment.
  • Strong experience engaging senior business stakeholders, defining measurable success criteria and translating model outputs into operational or commercial outcomes.
  • Demonstrated ability to mentor data scientists, improve engineering standards and influence technical direction without relying solely on formal authority.
  • Proven experience leading technical design, model review and delivery across multidisciplinary teams.

Technical Skills

Programming & Query Languages

Python (pandas, PySpark), SQL (Hive, Trino)

Big Data & ML Platforms

Databricks

Machine Learning

K-Means/clustering, XGBoost, CatBoost, LightGBM, classification, regression, anomaly detection etc

MLOps & DevOps

MLflow, CI/CD pipelines, Azure DevOps, model monitoring & drift detection

Visualisation & Reporting

Power BI, Databricks

Behavioural & Core Competencies

  • Strategic Thinking & Problem Solving - able to translate ambiguous business problems into structured analytical approaches.
  • Analytical & Innovative - rigorous, curious, and constantly looking for better ways to model and interpret data.
  • Communication & Business Storytelling - able to explain technical concepts and results in clear, business-friendly language for both technical and executive audiences.
  • Cross-Functional Collaboration - comfortable partnering with business, engineering and governance stakeholders.
  • Attention to Detail & Ownership - takes accountability for code quality, model accuracy and documentation.
  • Adaptability & Resilience - thrives in a fast-paced environment spanning multiple concurrent projects and priorities.
  • Technical Leadership & Judgement - sets direction, challenges weak assumptions and makes sound decisions under uncertainty.
  • Coaching & Capability Building - develops others through practical guidance, constructive review and meaningful delegation.
  • Commercial Acumen & Value Orientation - connects analytical work to measurable customer, revenue, cost, risk or experience outcomes.
  • Governance & Risk Ownership - anticipates model, data and operational risks and ensures appropriate controls are built into delivery.

About this listing

Verified live 1 day ago
Removed when the employer takes it down

No resume yet? Fix that first

Build a free, recruiter-ready resume in minutes, then tailor it to this job in one tap.

Create a free resume