Senior Data Scientist - Advanced AI Track

About the role

Singapore Airlines' in-house AI Center of Excellence builds and runs the company's shared AI platforms and delivers LLM and AI agent products to business units. This Senior Data Scientist role sits within that team and works closely with Data Engineering, Enterprise Architecture and business units.

Key responsibilities

• Build and operate shared AI platforms: a multi-provider LLM gateway, an agent orchestration layer, an evaluation and monitoring layer and a staff assistant.
• Deliver LLM and agent applications for business units, from prototype to live service with measurable results.
• Evaluate AI systems: build test sets, measure groundedness, accuracy, task success, safety, latency, reliability and drift, set release gates and rollback criteria, and manage model upgrades and retirements.
• For agentic workflows, evaluate tool-call accuracy, policy compliance, prevention of unauthorised actions and human escalation effectiveness.
• Balance cost and performance by routing between models, using batch processing and running models on company GPUs when data cannot leave the company.
• Design how agents retrieve and share business knowledge, with access control and data governance across SIA and Group companies.
• Apply traditional machine learning and optimization to fraud and anomaly detection, forecasting, resource optimization and document classification, deciding when a classic model beats an LLM.
• Act as the technical go-to person, work with business units to define problems and success measures, and plan and run AI projects end to end with clear status and risk reporting.

Required profile

• Degree in computer science, machine learning or a related field, or equivalent experience.
• 3+ years delivering ML or AI systems to production.
• Business communication, curiosity, logical thinking, can-do attitude and project management skills.
• Mentoring skills to guide junior data scientists.
• Experience taking an AI or analytics solution beyond launch and measuring its impact on revenue, cost, productivity, service quality or operations.

Required skills

• Strong Python and SQL.
• Traditional ML: gradient-boosted trees, regression, forecasting, clustering, anomaly detection, feature engineering and validation.
• LLM applications with commercial or open-weight models (Claude, GPT, Gemini, AWS Bedrock): prompts, retrieval, tool calling and multi-step agents.
• Structured evaluation and monitoring of LLM systems in production, including tool-call accuracy and policy compliance for agents.
• Responsible AI controls such as guardrails, groundedness checks and audit trails, plus data privacy and access control (e.g. PDPA).
• Secure cloud deployment (AWS, Azure or GCP) and cost management.