Senior Data Scientist: AI Forecasting in Production

UMATR; Abu Dhabi; On-site; AED 420,000 - 720,000; Full time

Job summary

UMATR in Abu Dhabi is seeking an experienced data scientist to build and deploy predictive models for risk assessment, forecasting and optimisation. You will develop AI-powered workflows and orchestrate LLM-driven processes.

You will own the full model lifecycle from experimentation to deployment, monitor performance and ensure guardrails. Strong Python, SQL and ML tooling are essential, plus experience with XGBoost/LightGBM and SHAP.

Qualifications

• 6+ years of experience in data science, AI/ML or quantitative modelling.
• Production-grade experience with forecasting, regression, classification, and optimization.
• Hands-on with XGBoost or LightGBM and SHAP for model interpretation.
• Strong Python (Pandas, Scikit-learn) and SQL skills.

Responsibilities

• Build and deploy predictive models for risk assessment and forecasting.
• Develop AI workflows with guardrails and oversight.
• Orchestrate LLM-powered workflows with structured and unstructured data.
• Own full model lifecycle from experimentation to deployment and monitoring.
• Establish evaluation, versioning and monitoring practices.
• Develop model-serving APIs for scalable production systems.
• Work with large datasets to identify patterns and opportunities.
• Build dashboards translating outputs into insights.
• Collaborate with engineering and senior business stakeholders.
• Present findings clearly to technical and non-technical audiences.

Skills

Forecasting

Regression

Classification

Optimization

XGBoost

LightGBM

Python (Pandas, Scikit-learn)

SQL

MLOps

LLMs

Communication

Education

Master's degree in Data Science

Tools

Python

SQL

Pandas

Scikit-learn

Job description

UMATR in Abu Dhabi is seeking an experienced data scientist to build and deploy predictive models for risk assessment, forecasting and optimisation. You will develop AI-powered workflows and orchestrate LLM-driven processes.

You will own the full model lifecycle from experimentation to deployment, monitor performance and ensure guardrails. Strong Python, SQL and ML tooling are essential, plus experience with XGBoost/LightGBM and SHAP.