Data Scientist

The Model Risk Management Department at Bank of Georgia is looking for a passionate, curious, and talented Data Scientist to join our growing team. You will have an opportunity to contribute to various innovative data-driven products and scalable intelligent systems. Members of the team are testing, validating, challenging and monitoring all the AI/ML models and solutions developed/used in the bank in order to mitigate model risk and create a safe environment for operations in different areas of business (credit risk, marketing, customer care, optimization of back-office operations, etc.). We are looking for people with hands-on experience in model development and/or validation who can independently identify, assess and mitigate model risks and take ownership of validation engagements from start to finish. If you are a resourceful, goal-oriented person who enjoys collaborating to deliver solutions, this may be an opportunity for you.  

Key Responsibilities:  

Responsibilities below are carried out with a high degree of independence; complex or high-impact matters are discussed with senior team members and the Head of Department as appropriate.  

• Independently lead model validations end-to-end, from scoping and planning to final reporting, assessing model performance, conceptual soundness, input data quality, and development methodology.  

• Design and carry out model outcomes analysis, including back-testing, sensitivity analysis, and stress testing of model behavior.  

• Evaluate new and existing models' overall fit and suitability with their intended use and purpose and constructively challenge model owners and developers on key design choices.  

• Develop a thorough understanding of the business processes, products, and decision-making context in which models are used, and assess whether model design, inputs, and outputs are appropriate for that context.  

• Identify models' key assumptions, limitations, and weaknesses, assess the effectiveness of compensating controls, rate the severity of findings, and formulate practical risk mitigation recommendations.  

• Develop benchmark and challenger models, selecting appropriate techniques to substantiate the assessment of models under review.  

• Prepare clear, well-structured validation reports and present findings and recommendations to model owners, developers, and business stakeholders.  

• Apply the Model Risk Policy and related standards consistently in day-to-day work and contribute to their review and enhancement.  

• Perform ongoing model performance monitoring, define appropriate metrics and thresholds, and follow up on the remediation of identified issues.  

• Contribute actively to model governance activities and drive the optimization of the validation and monitoring process through automation and reusable tools, incorporating regulatory changes and other initiatives.  

• Validate advanced AI/ML and Generative AI solutions (including LLM- and RAG-based applications) and contribute to developing the team's validation approaches for these model types.  

• Contribute to knowledge sharing within the team and support colleagues where needed.  

Preferred Qualifications:  

• Degree in Statistics, Mathematics, Data Science, Computer Science, or related fields.  

• 1–3 years of relevant working experience in model development, model validation, data science, or quantitative analysis.  

• Experience in the financial sector, particularly banking (e.g., credit risk, IFRS 9, or scoring models), is a plus.  

• Proven experience working with complex and large-scale structured datasets; experience with unstructured data (text, documents) is a plus.  

• Strong knowledge of statistical modeling and machine learning techniques, including model evaluation, explainability, and bias/fairness assessment.  

• Practical experience with deep learning, reinforcement learning, or natural language processing is an advantage.  

• Hands-on experience with Large Language Models (LLMs) and their practical applications, including prompt engineering, fine-tuning, or LLM evaluation, is an advantage.  

• Solid understanding of transformer architecture and attention mechanisms; practical experience with transformer-based models is an advantage.  

• Understanding of Retrieval-Augmented Generation (RAG) systems and their components, including vector databases, embedding models, and document retrieval pipelines, as well as approaches to evaluating them, is a plus.  

• Strong proficiency in Python and SQL, with solid knowledge of relational database concepts and data modeling; ability to write clean, reproducible, and well-documented code.  

• Experience with ML frameworks (TensorFlow, PyTorch, Scikit-Learn, etc.); familiarity with Big Data technologies (Hadoop/Hive, Spark, NoSQL, etc.) is a plus.  

• Experience with data visualization libraries (such as Matplotlib or Plotly) and tools (such as Microsoft Power BI or Tableau), including building dashboards and monitoring reports.  

Knowledge, Skills and Abilities:  

• Strong data storytelling skills and the ability to communicate complex analytical findings clearly and persuasively to both technical and non-technical audiences, including senior stakeholders.  

• Strong analytical and problem-solving skills.  

• Ability to analyze and understand business processes and translate business context into model risk considerations.  

• Critical thinking, professional skepticism, and attention to detail.  

• Team player and self-starter with a strong sense of ownership, able to manage multiple validations and priorities in parallel and deliver to deadlines.  

• Ability to troubleshoot issues independently and exercise sound judgment on when to escalate.  

• Proficiency in writing clearly and concisely in both Georgian and English. 

Interested candidates, please fill in the information, attach your CV and submit by clicking “apply for position now”. Deadline is November 18 2026.