The AI & Data Enterprise Architect is responsible for defining and governing the Group-wide AI target state spanning GenAI, Agentic AI, ML, and decisioning systems while embedding robust data architecture foundations to ensure AI is trustworthy, explainable, and compliant. This role drives innovation and adoption of AI use cases, aligning digital technology solutions with evolving business needs and changing market dynamics.
As a key focus, the position will define and steward the Bank's next-generation AI & Data reference architecture, leading efforts in modernization, modularization, and seamless integration to ensure solutions are secure, future-ready, and interoperable across ASEAN markets.
We seek a senior architect with proven hands-on GenAI production experience (3+ years since 2022 adoption), combined with 5+ years in applied AI/ML, built on enterprise-grade platforms such as Microsoft Azure OpenAI and Copilot ecosystems. Exposure to AI governance, LLMOps, RAG, and model safety is required. The ideal candidate will also bring strong external presence, with experience presenting at industry events, forums, and conferences, and excellent communication skills to influence both internal and external stakeholders.
Key Responsibilities:
1. AI Strategy & Target-State Architecture
- Define the Group AI target state architecture, spanning GenAI, Agentic AI, ML, and decisioning systems.
- Establish reference architectures for high-value use cases (customer service, underwriting/credit, fraud & FC, KYC, productivity agents, developer copilots).
- Guide platform selection across cloud/on-prem, with emphasis on Microsoft AI environments (Azure OpenAI Service, Copilot, Fabric, Purview, Responsible AI tooling).
2. Agentic AI & LLMOps Platforms
- Lead adoption of Agentic AI patterns (multi-agent orchestration, planning, tool-use, long-horizon reasoning, dynamic workflows).
- Institutionalize LLMOps/MLOps for agent-based systems: lifecycle management, evaluations, observability, safety nets, fallback strategies.
- Standardise orchestration frameworks (LangChain, Semantic Kernel, Azure AI Studio) for Agentic AI + RAG.
- Implement cost, safety, and performance monitoring for autonomous AI systems.
3. Data & Model Governance
- Embed AI governance aligned to NIST AI RMF, ISO/IEC 42001 (AI management systems), and ISO/IEC 23894 (AI risk management).
- Apply MAS TRM, BNM RMiT, OJK MRTI, PDPA/GDPR, and MAS FEAT principles for responsible AI in finance.
- Operationalize Microsoft Responsible AI standards and Azure AI compliance features (AI dashboards, fairness/bias detection, interpretability).
- Govern models via registries, lineage, model cards, approval gates, red-teaming, bias/hallucination tests, and periodic revalidation.
4. Architecture & Engineering Enablement
- Publish secure RAG + Agentic AI patterns (citation, source attribution, guardrails, hybrid search, freshness).
- Implement defence-in-depth for GenAI/Agentic AI (prompt-injection shielding, content filtering, PII redaction, isolation).
- Guide model selection/tuning (few-shot, fine-tuning, adapters, LoRA) and GPU/accelerator trade-offs.
- Integrate with Microsoft Fabric for analytics & data pipelines, Purview for governance, and Synapse/Data Lake for training and inference.
5. Risk, Security & Controls
- Integrate AI risks into enterprise risk frameworks; ensure Architecture Review Board reviews cover Agentic AI autonomy risks, data risks, and model safety.
- Define incident playbooks (prompt-injection, rogue agents, data exfiltration, model compromise).
- Ensure comprehensive auditability (logs, evaluations, reproducibility, DPIAs, red-team reports) across Microsoft AI and multi-cloud deployments.
6. KPI / Success Measures
- Develop and maintain the AI and Data target state architecture, acting as the trusted Enterprise Architect for business and technology leaders to align all digital initiatives with Group strategy.
- % of AI initiatives delivered on Group AI/Agentic AI reference architectures and governance standards.
- % of ARB approvals without high-risk exceptions.
- Business uplift attributable to AI (NPS, fraud catch rate, cost-to-serve).
- Cost efficiency across AI/LLM workloads (tokens, GPU hours).
Qualifications and Experience:
- 15+ years in enterprise/solution architecture, including 5+ years applied AI/ML and 3+ years hands-on GenAI/LLM since 2022.
- Proven design of Agentic AI systems and orchestration (LangChain, Semantic Kernel, Azure AI Studio).
- Experience operationalizing AI governance (NIST, ISO/IEC, MAS FEAT, Microsoft Responsible AI) and preparing for EU AI Act.
- Strong Microsoft ecosystem experience: Azure OpenAI, Copilot, Fabric, Purview, Synapse, Cognitive Services.
- Bachelor's in CS/Eng/IS. Microsoft AI/Cloud certifications highly desirable.
Desired Skills and Competencies:
- Expert in GenAI + Agentic AI (reasoning, multi-agent orchestration, autonomous planning).
- Proficient in LLMOps/MLOps toolchains (Azure ML, MLflow, LangChain, Semantic Kernel, vector DBs, eval frameworks).
- Strong security & privacy mindset (data sovereignty, PII safeguards, secure enclaves).
- Leadership presence to influence CxO, risk, regulators; strong vendor management.
- Passion for safe, responsible, high-impact AI, with ability to translate into measurable outcomes.
