Senior Manager - AI Safety and Evaluation Engineering

About the role

Join Abu Dhabi Commercial Bank (ADCB) as a Senior Manager leading AI Safety and Evaluation Engineering. You will shape the reliability, safety and security of the bank’s AI solutions, ensuring they meet rigorous risk and compliance standards while driving responsible AI adoption across the organization.

Key responsibilities

• Design, implement and scale technical measures to guarantee the safety and security of predictive ML and Generative AI models.
• Define standardized evaluation rubrics and scoring systems for AI performance, reliability and safety across different GenAI archetypes.
• Create comprehensive test cases covering deterministic logic, probabilistic outputs, edge‑case and adversarial scenarios.
• Implement Explainable AI (XAI) techniques such as Shapley Values and Integrated Gradients, and deploy bias detection metrics to ensure model transparency and regulatory compliance.
• Collaborate with MLOps and LLMOps teams to embed evaluation and monitoring capabilities into the production pipeline.
• Design an integrated observability architecture with telemetry pipelines that capture inputs, outputs and embeddings while preserving PII and data‑privacy standards.
• Establish “Golden Signals” (latency, drift, hallucination rates, semantic similarity) and automated circuit‑breakers to protect model health.
• Serve as the technical liaison to second‑line risk functions, aligning AI deployments with the bank’s AI Governance, Ethics and Responsible AI frameworks.

Required profile

• Proven experience in AI safety, evaluation or risk engineering for enterprise‑scale ML and Generative AI systems.
• Deep understanding of AI Governance, AI Ethics and Responsible AI principles.
• Strong analytical mindset with the ability to translate technical findings into actionable risk mitigations.
• Excellent communication skills to work effectively with risk, compliance and engineering stakeholders.

Required skills

• Machine Learning and Generative AI
• Explainable AI (XAI) techniques – Shapley Values, Integrated Gradients
• Bias detection and fairness metrics
• MLOps and LLMOps platforms
• Telemetry, observability and monitoring pipelines
• Data privacy and PII handling
• Model performance metric frameworks (latency, drift, hallucination rates, semantic similarity)