New opportunity

Lead AI Platform Engineer #AIDA

Singtel Group · Singapore

About this role

Lead the implementation and operation of shared Agentic AI Platform Services for Singtel’s Central AI Kitchen, running on RE:AI infrastructure using Red Hat OpenShift or an equivalent container platform. The role develops reusable, governed platform capabilities that let application teams build, onboard and operate enterprise AI agents, while improving their security, scalability, interoperability, resilience and reliability.

Responsibilities

Implement and manage the lifecycle of assigned agent services. Design and operate MCP servers, MCP/API gateways, and agent and tool gateways that provide governed connectivity to enterprise tools, APIs and data services, including access control, routing and policy enforcement.
Implement agent identity and authentication with the NCS development team to support trusted agent-to-tool, service-to-service and agent-to-agent interactions. Operate an enterprise agent registry for discovery, metadata, ownership, lifecycle status and governance.
Establish reusable onboarding standards for agents, MCP servers, tools and APIs. Apply authentication, authorization, secrets management and least-privilege controls; integrate services with enterprise IAM, API management, SIEM, logging and cybersecurity systems. Support agent-to-agent interoperability and emerging communication standards where applicable.
Lead engineers across Singapore and India through technical direction, planning, task allocation, mentoring and delivery oversight. Coordinate internal staff, offshore engineers, vendors and other platform teams; establish coding, documentation, design review, testing, knowledge-sharing and handover practices. Collaborate with AI architects, infrastructure, cybersecurity and application teams.

Requirements

A bachelor’s degree in Computer Science, Engineering or a related field; at least 8 years of experience in platform engineering, cloud-native infrastructure or data/AI platform roles; experience implementing platforms on Kubernetes or OpenShift; experience designing or integrating enterprise APIs, gateways, identity platforms or distributed application services; and experience leading platform implementation and coordinating engineers across teams or locations.

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

Skills for this role

Agentic AIModel Context Protocol (MCP)MCP serversAPI gatewaysAgent identityAgent registryRed Hat OpenShiftKubernetesPlatform engineeringCloud-native infrastructureEnterprise APIsAuthenticationAuthorizationIdentity and access management (IAM)Secrets managementSIEMAgent-to-agent interoperabilityCybersecurityTechnical leadershipMentoringTesting

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

Deliver and operate shared agent services
The role leads implementation and lifecycle management of Central AI Kitchen services on RE:AI infrastructure using OpenShift or an equivalent container platform, with attention to scalability, resilience and reliability.
Build governed agent connectivity
Design and operate MCP servers, MCP/API gateways, and agent and tool gateways that connect agents to enterprise tools, APIs and data services while enforcing access and routing policies.
Establish trusted identities and controls
Implement agent identity and authentication with the NCS development team; apply authorization, secrets management and least-privilege controls, and integrate with enterprise security services.
Make agent onboarding reusable and governable
Establish onboarding patterns for agents, MCP servers, tools and APIs, and operate a registry covering discovery, metadata, ownership and lifecycle status.
Lead delivery across locations and teams
Provide technical direction, planning and oversight for engineers in Singapore and India, while coordinating vendors and platform teams and establishing review, testing, documentation and handover practices.

Likely challenges to think through

Balance interoperability with controlled access
Preparation hypothesis: standardized agent-to-tool and agent-to-agent interactions may need to accommodate different enterprise systems without weakening policy enforcement. Prepare to explain design trade-offs for gateway boundaries, identity, routing and least-privilege access.
Balance reuse with service-specific needs
Preparation hypothesis: common onboarding patterns and registry rules may need to serve varied agents, tools and APIs without making onboarding impractical. Prepare an example of how you would define a reusable baseline while handling justified exceptions and lifecycle changes.
Coordinate delivery without losing operational quality
Preparation hypothesis: work split across Singapore, India, vendors and other platform teams could make ownership and handovers harder while services must remain reliable. Prepare to describe how you would allocate responsibility and use design reviews, testing, documentation and delivery tracking to manage that trade-off.

Practice your answers

Technical practice

How would you deploy and operate shared agent services on OpenShift or an equivalent container platform?

What to coverPreparation suggestion: Explain lifecycle management, deployment boundaries, configuration and operational ownership. Discuss how you would evaluate scalability and reliability for services shared by application teams.

How would you design an MCP server that connects agents to enterprise tools or data services?

What to coverPreparation suggestion: Cover tool interfaces, request validation and governed access to downstream services. Explain how you would test the server and handle downstream failures without exposing more access than needed.

What responsibilities would you place in MCP, API, agent and tool gateways?

What to coverPreparation suggestion: Compare where to enforce access policies, route requests and integrate with enterprise services. Address the trade-off between reusable central controls and flexibility for individual services.

How would you establish trusted identities for agent-to-tool and agent-to-agent interactions?

What to coverPreparation suggestion: Distinguish agent identity, authentication and authorization across interaction types, including service-to-service calls. Explain how you would coordinate interfaces and trust boundaries with the development team responsible for related identity capabilities.

How would you design an enterprise agent registry?

What to coverPreparation suggestion: Describe metadata for discovery, ownership and lifecycle status, and how updates would be governed. Explain how consumers could use the registry without treating an entry as proof that an agent is authorized to access a tool.

What reusable onboarding standard would you establish for agents, MCP servers, tools and APIs?

What to coverPreparation suggestion: Outline the information, access controls, testing and handover needed before a component is operated on a shared platform. Discuss how to make the pattern repeatable while accommodating different integrations.

How would you enforce least-privilege tool access and manage secrets for AI agents?

What to coverPreparation suggestion: Cover authorization scope, credential handling and controls that prevent an agent from invoking tools beyond its approved permissions. Explain how you would assess changes to access without relying on broad shared credentials.

How would you integrate agent services with IAM, API management, logging and SIEM?

What to coverPreparation suggestion: Trace an access request and its operational or security events across those enterprise services. Discuss what evidence operators need for troubleshooting and security review, while limiting sensitive data in logs.

How would you approach agent-to-agent interoperability as communication standards evolve?

What to coverPreparation suggestion: Explain how you would define interaction boundaries and assess a standard before adopting it. Consider compatibility, identity and policy enforcement across agents rather than assuming one protocol solves every integration.

How would you evaluate the resilience and reliability of a shared agent platform service?

What to coverPreparation suggestion: Discuss failure scenarios, testing and operational signals for gateways or agent services. Explain trade-offs between availability, safe failure of governed access and the complexity of recovery.

Behavioural practice

Tell me about a time you led implementation across engineers in different locations.

What to coverPreparation suggestion: Use a real situation, your approach to planning, task allocation and delivery tracking, and the result. Clarify your own technical decisions and how you handled coordination across locations.

Describe a time you resolved a disagreement between platform delivery and security needs.

What to coverPreparation suggestion: Give the situation, the stakeholders' concerns, the actions you took to reach a decision and the result. Keep the example grounded in work you actually did rather than claiming experience with this employer's systems.

Tell me about a time you improved engineering practices or mentored a team.

What to coverPreparation suggestion: Identify a specific gap in reviews, testing, documentation or knowledge sharing; explain your actions and the result. Distinguish what you introduced from what the team adopted.

Describe an operational problem in a shared platform service that you helped resolve.

What to coverPreparation suggestion: Set out the situation, how you investigated and coordinated the response, and the outcome. Include what changed afterward to improve reliability or handover, if that was part of your real example.

Tell me about a delivery dependency involving another team or vendor.

What to coverPreparation suggestion: Explain the situation, how you clarified responsibilities and managed the dependency, and the result. Focus on actions you controlled and how you kept affected teams informed.

Sharpen your knowledge

Topics to prepare

Shared agent platform architecture
Posting: Central AI Kitchen runs shared Agentic AI Platform Services on RE:AI infrastructure using OpenShift or an equivalent container platform. Revision exercise: sketch how application teams would onboard and operate agents through shared services, marking platform ownership boundaries and integration points.
Governed MCP and gateway design
Posting: MCP servers and MCP, agent and tool gateways connect agents to enterprise tools, APIs and data services with access control, routing and policy enforcement. Revision exercise: trace one agent-to-tool request through these components and explain where each control applies.
Agent identity and least privilege
Posting: trusted agent-to-tool, service-to-service and agent-to-agent interactions require identity, authentication, authorization and secrets management. Revision exercise: map the identities and permissions in one interaction, then identify where a secret or overly broad tool permission would create risk.
Registry and agent interoperability
Posting: the agent registry supports discovery, metadata, ownership, lifecycle status and governance; agent-to-agent interoperability is supported where applicable. Revision exercise: outline a registry record and explain how its status and ownership would affect discovery and use. Treat any specific communication standard as a discussion point, not a stated requirement.
Reliability and engineering governance
Posting: the platform is expected to be secure, scalable, resilient and reliable, and the lead establishes design reviews, testing, documentation and handovers across Singapore and India. Revision exercise: review a proposed platform change and identify its operational risks, tests, review participants and handover information.

Skills to revise

Deploy a platform component
Posting: platform implementation experience on Kubernetes or OpenShift is required. Suggested practice: prepare a deployment and operations walkthrough for one assigned agent service, covering configuration, scaling, failure recovery and an update. The walkthrough is preparation, not a prescribed employer procedure.
Specify an MCP integration
Posting: the role designs, implements and operates MCP servers and gateway capabilities. Suggested practice: draft a small MCP server-to-gateway integration specification for an enterprise API, including routing, permitted access and a test for denied access.
Apply identity and security controls
Posting: the role integrates with IAM, API management, secrets management, SIEM and logging services, collaborating with the NCS development team on agent identity. Suggested practice: write an integration checklist for authentication, authorization, secret handling and security-event logging in an agent-to-tool flow.
Operate a registry lifecycle
Posting: the role implements and operates an enterprise agent registry. Suggested practice: draft example metadata for an agent and walk through registration, an ownership update, a lifecycle-status change and discovery, stating what must be checked at each step.
Lead a cross-location delivery review
Posting: the lead plans and allocates work, tracks delivery and sets testing, documentation, design-review and handover practices for engineers across Singapore and India. Suggested practice: split a gateway change into reviewable tasks, define acceptance tests and prepare a concise handover for the operating team.

Your preparation checklist

Map the stated requirements to evidence
Prepare a one-page evidence map for the listed degree, 8+ years in relevant platform roles, Kubernetes or OpenShift implementation, enterprise APIs or identity integration, and technical leadership. Note the specific example you would discuss for each.
Build a platform implementation walkthrough
Prepare a diagram and concise account of a platform component you implemented: architecture, deployment, your decisions, operational responsibilities and outcomes. Relate it to the role’s shared services on RE:AI infrastructure using OpenShift or an equivalent container platform without implying you have worked on Singtel’s platform.
Practice an MCP gateway design exercise
As interview preparation, sketch how an MCP server and MCP/API gateway could connect agents to enterprise tools and data. Explain routing, access control, policy enforcement, failure handling and what you would clarify before implementation.
Trace identity and governance end to end
Prepare a sample agent-to-tool interaction showing authentication, authorization, secrets handling and least-privilege access. Add a proposed registry record covering discovery metadata, ownership and lifecycle status; the role calls for these capabilities, while your example is a preparation exercise.
Outline onboarding and operations
Draft a short checklist for onboarding an agent, MCP server or tool, including testing, logging and integration points with IAM, API management and SIEM. Be ready to explain how you would monitor reliability and manage a service through its lifecycle.
Document a cross-location leadership example
Prepare one specific account of planning work, assigning tasks, reviewing designs, mentoring engineers and coordinating handovers across teams or locations. Identify your contribution, the delivery challenge and the result.
Bring questions about delivery boundaries
Write questions for the employer about ownership of Central AI Kitchen services, collaboration with the NCS development team on agent identity, integration with security teams, and how Singapore and India engineers coordinate delivery.

Availability last checked: 4 Nov 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…