About the Role

We are seeking a passionate and technically strong Principal AI Architect to join Kuok Group’s AI Centre of Excellence. This person will own the architecture, engineering standards, and technical infrastructure that every AI use case across the Group runs on — from the LLM gateway and RAG data foundation through to CI/CD pipelines, model observability, and governance controls.

The successful candidate will lead the AI Platform cluster directly, set technical standards for the Applied AI cluster, and serve as the engineering authority whose review and approval gates every production deployment — operating as architect, reviewer, and standards-setter to enable the rest of the team to build at speed without compromising on quality or security.

This is a platform-first role, focused on building the durable technical foundation that the Group’s AI capability compounds on overtime.
.

Key Responsibilities

Architecture & Technical Standards

• Design and own the end-to-end AI platform architecture — LLM gateway, RAG layer, vector + knowledge graph foundation, model serving, observability, and eval infrastructure
• Set and enforce engineering standards across both the AI Platform and Applied AI clusters: code review criteria, security baselines, deployment gates, and documentation requirements
• Own the technical roadmap for the Group's broader AI infrastructure stack

Platform Cluster Leadership

• Manage the AI Platform team cluster directly: AI Platform Engineer, AI Data Architect, and LLM Ops / MLOps Engineer
• Provide technical direction and functional oversight to the AI Data Architect on AI data foundation design — embedding strategy, vector DB design, knowledge graph ingestion approach
• Coordinate with the Applied AI cluster on technical requirements for use case delivery and review Applied AI Engineers' architecturally significant decisions
• Own offshore and vendor technical oversight: standing weekly tech review, templated acceptance criteria, and code review consistency for all externally delivered work

Platform Build & Infrastructure

• Partner with the AI Platform Engineer on CI/CD pipeline design, environment provisioning, infra-as-code, and the deployment platform (staging, production, rollback)
• Oversee the model serving and gateway layer including API access, rate limits, and model switching controls
• Ensure LLM observability (LangSmith and similar), cost telemetry, and token controls are in place and monitored as a first-class concern
• Establish and maintain the eval pipeline infrastructure in partnership with the LLM Ops /MLOps Engineer; make evals a gated handover criterion for every production use case

AI Data Foundation

• Work with the AI Data Architect to design the hybrid vector + knowledge graph layer that underpins RAG across all BUs — the enterprise 'semantic memory'
• Set the strategic direction for embedding pipelines, chunking strategies, and vector DB design, with the AI Data Architect owning day-to-day execution

Governance, Security & Risk

• Partner with the AI Governance & Compliance Lead on security architecture, red team drill cadence, and data privacy protocols across all deployed agents and workflows
• Provide technical inputs to AI Governance & Compliance Lead

Stakeholder Engagement

• Represent the technical platform view in Steerco updates, architecture reviews, and vendor selection decisions
• Translate platform architecture constraints and investments into business-legible language for non-technical senior stakeholders
• Evaluate and negotiate with vendors on technology stack decisions — model providers, observability tooling, infra providers

.

Requirements

Must-Have

• Proven architecture experience in AI / ML systems — you have designed and shipped production systems that include LLM-based components, not just traditional ML pipelines
• Strong engineering fundamentals: proficiency in Python, cloud infrastructure (Azure preferred), and DevOps practices including CI/CD, IaC, and environment management
• Hands-on experience with LLM infrastructure: model gateway design, prompt versioning, observability tooling (LangSmith or equivalent), and cost controls
• Demonstrated experience with RAG architecture — chunking strategies, vector DB design, retrieval optimisation, re-ranking, and knowledge graphs
• Agentic AI framework knowledge: LangGraph, CrewAI, or equivalent — and an understanding of the operational differences between stateless and stateful agent patterns
• Engineering leadership track record: you have set technical standards, run code reviews, and led a team to deliver to a high bar
• Strong communication skills — able to write clearly, produce architecture decision records, and present to technical and non-technical audiences

.

Strong Advantage

• Experience with knowledge graph design and ontology modelling (Neo4j or equivalent)
• MLOps or LLMOps background: model versioning, A/B testing for model outputs, eval pipeline design, LLM-as-judge patterns
• Security architecture experience relevant to AI systems: threat modelling for agentic workflows, data privacy design, red teaming
• Vendor management experience — MSA negotiation, offshore technical oversight, acceptance criteria design
• Familiarity with enterprise conglomerate environments: multiple BUs, federated data, diverse system landscapes