Global Life Insurer - Asia MLOps/LLMOps Engineer (Regional)

Argyll Scott; Hong Kong; On-site; HKD 900,000 - 1,500,000; Full time

Job summary

Argyll Scott is seeking an ML/AI Engineer to join its Regional AI Centre of Excellence in Hong Kong. You will own end-to-end ML production pipelines from data ingestion to monitored endpoints, focusing on lakehouse architectures and LLM services.

The role emphasizes MLOps, model registries, feature stores, automated retraining and governance, with deployment on AWS (SageMaker/Bedrock) and potential GCP support. 8+ years in ML/data engineering required, including 4+ years in production.

Qualifications

• 8+ years in ML, data or platform engineering, including 4+ years operating ML systems in production.
• Experience deploying ML workloads on cloud platforms (AWS, GCP) with services like SageMaker/Bedrock.

Responsibilities

• Build reproducible training pipelines from data extraction to model promotion.
• Integrate ML workloads into medallion lakehouse architecture with data-quality controls.
• Operate feature stores and model registries with versioning and deployment records.
• Implement automated retraining, shadow and canary deployments, monitoring and rollback.
• Extend lifecycle controls to prompts, retrieval indexes and LLM evaluation.
• Develop model routing and gateway capabilities to manage latency and cost.

Skills

Python

SQL

Spark / PySpark

Tools

Iceberg

Delta Lake

MLflow

Docker

Kubernetes

SageMaker

Bedrock

Job description

ML/AI Engineer — MLOps & LLMOps

Regional AI Centre of Excellence | Global Life Insurer
Location: Hong Kong; candidates based in the Philippines or India may also be considered
Experience: 8+ years in ML, data or platform engineering, including 4+ years operating ML systems in production

The opportunity

We are partnered with a global life insurer to build out its regional AI Centre of Excellence, supporting transformation across its Asian businesses. Bringing together engineering, data science, product and business expertise, the Centre connects advances in machine learning, generative AI and agentic systems with opportunities across distribution, customer engagement and operations.

As an ML/AI Engineer, you will own the engineering lifecycle that takes data, models and AI services into reliable production. Your remit will span lakehouse pipelines, training and feature engineering, model deployment, LLM operations and automated governance. This is an opportunity to build shared capabilities that support multiple use cases and markets.

What you will do

• Build reproducible training pipelines covering data extraction, feature computation, tuning, evaluation, model registration and controlled promotion.
• Work with platform teams to integrate ML workloads into medallion lakehouse architecture, implementing data-quality controls and efficient incremental processing.
• Operate feature stores and model registries, with clear versioning, lineage, approval states and deployment records.
• Implement automated retraining, shadow and canary deployments, monitoring and rollback for batch and real-time services.
• Extend lifecycle controls to prompts, retrieval indexes and agent configurations, including RAG pipelines, vector database operations and LLM evaluation.
• Develop model routing and gateway capabilities that manage latency, cost, rate limits and provider failover.
• Deploy primarily on AWS, using services including SageMaker and Bedrock, with GCP support where appropriate.
• Automate asset inventories, model metadata and compliance evidence flows into the AI governance platform.

What you will bring

• Proven ownership of production pipelines from raw data ingestion through to monitored endpoints.
• Strong Python and SQL skills, with distributed processing experience using Spark, PySpark or equivalent.
• Hands‑on experience with medallion lakehouse architecture and open table formats such as Iceberg or Delta Lake.
• Strong MLOps tooling experience across experiment tracking, model registries, feature stores, orchestration and monitoring.
• Experience with MLflow, Docker/Kubernetes and at least two cloud‑native platforms.
• Deep AWS knowledge across access management, networking, data services, serverless and containers.
• Experience integrating governance tooling or automating compliance evidence collection.

Preferred: GCP/Vertex AI; LLMOps and RAG operations; streaming architectures; data-quality and catalogue tooling; feature-store implementation; GPU cost optimisation; and infrastructure as code at scale.

Why this role

You will help establish the production foundations of a regional AI capability, combining deep engineering ownership with exposure to modern ML, LLM and agent technologies in a complex enterprise environment.