New opportunity

Lead, AI & Agent Operations #AIDA

Singtel Group · Singapore

About this role

The Lead, AI & Agent Operations #AIDA will oversee the Day 2 AI and Agent Operations function at Singtel, responsible for the production AI estate including GenAI agents and classical ML models. This role serves as the senior operational point of accountability, ensuring reliability, safety, and quality across the organization's AI systems. Responsibilities include managing the operating model from build to production, overseeing handover gates, 24x7 monitoring, incident response, and evaluation. The lead will manage distributed teams in Singapore and India, define cross-border operational workflows, and hold gate authority for AI deployments. Key duties involve governing deployment and change discipline, managing service-level objectives and KPIs, and reporting AI reliability to the Chief AI Officer and the board. Required qualifications include a degree in Computer Science, Engineering, Data Science, or a related field, and at least 10 years of experience in technology with a focus on production operations, SRE, or ML/AI functions. Candidates must demonstrate a track record of operating or governing ML/AI systems in production, experience leading remote or cross-border teams, and expertise in incident management and quality governance. Strong stakeholder management and executive communication skills are essential, along with an understanding of LLM/agent and traditional ML lifecycles and their failure modes. Knowledge of telecommunications consumer, enterprise (B2B), and network operations is highly relevant.

For full details about this role, click “Apply on company site” to view the employer’s original posting.

Skills for this role

AI OperationsAgent OperationsMachine Learning OperationsSREGenerative AILLMObservabilityIncident ManagementService-Level ObjectivesError BudgetsTeam LeadershipStakeholder ManagementCross-border Team Management

Checking your profile…

Preparation suggestions, not the employer’s actual interview questions. Always check the original posting for current requirements.

Understand the role

What matters most

Establishing Day 2 AI Operational Governance
The primary mandate is to define and enforce the end-to-end operating model for production AI. This requires creating robust handover gates between build and operations teams, implementing 24x7 monitoring, and establishing clear incident response protocols to ensure the reliability and safety of both GenAI agents and classical ML models.
Managing Cross-Border Team Synergy
You must lead and integrate distributed teams across Singapore and India. This involves defining operational workflows that bridge time zones, ensuring consistent quality standards, and maintaining team wellbeing while managing a unified, high-stakes production environment.
Defining AI Reliability and Performance Metrics
You are responsible for setting and reporting on Service-Level Objectives (SLOs), error budgets, and KPIs. This includes translating complex AI performance data into actionable insights for the Chief AI Officer and the board, ensuring that AI value and reliability are clearly communicated and aligned with business objectives.
Enforcing Quality and Deployment Gating
As the senior point of accountability, you hold go/no-go authority for AI deployments. You must maintain the integrity of evaluation processes, ensuring that deployment speed does not compromise safety or quality, and governing change discipline across the AI estate.

Likely challenges to think through

Balancing Delivery Speed with Rigorous Quality Gates
There is an inherent tension between the pressure to deploy new AI capabilities and the requirement to maintain strict quality and safety standards. Prepare to discuss how you have historically maintained independence and enforced 'stop-ship' decisions when models or agents failed to meet reliability thresholds.
Managing Non-Deterministic AI Failure Modes
Unlike traditional software, GenAI agents exhibit non-deterministic behaviors. You will face the challenge of designing observability and incident management systems that can detect and remediate 'soft' failures, such as hallucinations or performance degradation, rather than just binary system outages.
Aligning AI Operations with Complex Telco Infrastructure
The role requires deep integration with consumer, enterprise, and network operations. The challenge lies in understanding how AI failures impact these specific domains and designing fallback mechanisms that protect the customer experience during critical incidents.

Practice your answers

Technical practice

Defining AI Service-Level Objectives (SLOs)

What to coverExplain how you would define and measure SLOs for a production GenAI agent compared to a traditional ML model. Discuss the trade-offs between latency, accuracy, and cost, and how you would manage error budgets when performance degrades.

AI Deployment and Gating Strategy

What to coverDescribe your approach to establishing 'handover gates' between the build team and production operations. Detail the criteria you would use for a go/no-go decision to ensure model safety and reliability before deployment.

Observability for GenAI Agents

What to coverWhat specific metrics and observability patterns are essential for monitoring LLM-based agents in production? Address how you would detect 'hallucinations' or drift in real-time and trigger automated incident responses.

Incident Management for AI Systems

What to coverOutline your process for handling a critical failure in a customer-facing AI agent. Discuss the importance of rollback strategies, fallback mechanisms to human agents, and how you would conduct a post-incident review.

Managing AI Failure Modes

What to coverCompare the failure modes of traditional ML models versus LLM-based agents. Suggest how you would design operational workflows to mitigate these distinct risks, such as data drift for ML or prompt injection/safety issues for LLMs.

Quality Governance and Integrity

What to coverHow do you maintain objective quality standards when the same team is responsible for both building and evaluating AI agents? Suggest methods to ensure evaluation remains evidence-based despite delivery pressures.

Scaling AI Operations

What to coverAs the AI estate grows, how would you evolve your operational tooling and headcount strategy? Discuss the balance between manual oversight and automated governance to maintain reliability at scale.

Telco-Specific AI Applications

What to coverGiven the context of telecommunications, how would you tailor AI reliability strategies for network operations versus consumer-facing B2B services? Suggest how you would prioritize incident response based on customer impact.

Behavioural practice

Leading Cross-Border Teams

What to coverDescribe a time you managed a distributed team across different time zones and cultures. Focus on how you ensured clear communication, ownership, and team wellbeing while maintaining operational consistency.

Executive Stakeholder Management

What to coverProvide an example of how you communicated a complex technical failure or a significant operational risk to senior leadership. How did you translate technical metrics into business impact to secure support for remediation?

Maintaining Integrity Under Pressure

What to coverShare a situation where you had to make a difficult decision to halt a deployment or delay a release due to quality concerns. How did you handle the pushback from stakeholders while maintaining the integrity of the production environment?

Decision Making with Incomplete Information

What to coverDescribe a high-stakes incident where you had to make a critical decision with limited data. What was your process for assessing the situation, and what was the outcome of your action?

Sharpen your knowledge

Topics to prepare

AI and Agent Lifecycle Governance
Focus on the transition from Day 1 (build) to Day 2 (operations). Review frameworks for handover gates, deployment discipline, and regression testing. Exercise: Draft a hypothetical 'Go/No-Go' checklist for a GenAI agent deployment, specifically identifying criteria for safety, reliability, and performance that would trigger a rollback.
SRE Principles for AI Systems
Understand how to apply Site Reliability Engineering (SRE) to non-deterministic AI models. Focus on defining Service-Level Objectives (SLOs) and Error Budgets for AI. Exercise: Define three specific KPIs for an LLM-based customer service agent, such as hallucination rate or latency, and explain how you would manage an 'error budget' if these thresholds are breached.
Incident Management in AI Environments
Study incident response protocols specifically for AI, including fallback mechanisms to human agents or traditional systems. Exercise: Create a communication plan for a major AI system failure, detailing the escalation path to the Chief AI Officer and the steps for coordinating remediation across distributed teams.

Skills to revise

Observability and Evaluation Strategy
Develop expertise in monitoring both classical ML models and GenAI agents. Exercise: Outline a monitoring dashboard strategy that tracks both technical performance (latency, throughput) and model-specific quality metrics (accuracy, drift, safety guardrails).
Cross-Border Team Leadership
Focus on managing distributed operations between Singapore and India. Exercise: Propose a workflow for 24/7 coverage that ensures clear handoffs, maintains team wellbeing, and prevents silos between the two locations.
Executive Stakeholder Communication
Practice translating complex AI reliability data into business-relevant insights for board-level reporting. Exercise: Prepare a summary report template that presents AI reliability metrics, current incident status, and the strategic roadmap for the AI estate in a format suitable for a Chief AI Officer.

Your preparation checklist

Map Day 2 AI Operational Frameworks
Prepare a conceptual framework for transitioning AI models from build to production. Document how you would structure handover gates, define service-level objectives (SLOs), and manage error budgets specifically for GenAI agents versus classical ML models.
Develop an Incident Response Playbook for AI
Draft a high-level incident management strategy for AI-driven customer-facing systems. Include escalation paths, rollback procedures, and criteria for triggering human-in-the-loop fallback mechanisms during model failures or hallucinations.
Prepare Cross-Border Leadership Case Studies
Create two specific examples demonstrating how you have managed distributed teams across different time zones (e.g., Singapore and India). Focus on how you maintained team cohesion, standardized operational workflows, and ensured consistent quality governance.
Formulate Executive Reporting Metrics
Outline a dashboard or reporting structure suitable for a Chief AI Officer or Board level. Identify the key performance indicators (KPIs) you would track to demonstrate AI reliability, cost efficiency, and business value in a telecommunications context.
Draft Strategic Questions for Leadership
Prepare thoughtful questions regarding the current maturity of Singtel's AI estate, the existing relationship between Day 1 build teams and Day 2 operations, and the specific regulatory or safety constraints unique to their B2B and network operations.

Availability last checked: 18 Oct 2026. Status can change after a check; confirm on the employer’s website. Report an issue · Our checking policy

KEEP EXPLORING

Similar AI jobs

Related skills and specializations in Singapore. Matched to this role, not your profile.

Explore more jobs

Finding similar opportunities…