GenAI Data Platform Engineer

UOB KAY HIAN PRIVATE LIMITED; Singapore; Hybrid; SGD 120,000 - 190,000; Full time

Job summary

UOB KAY HIAN PRIVATE LIMITED is seeking a hands-on data engineer to own the enterprise Data Lake and AI Platform. You will design and maintain scalable data pipelines using AWS Glue, S3, Redshift, and related cloud-native tools while ensuring data quality, governance, and security.

The role emphasizes building AI agents, GenAI-powered solutions, and RAG architectures to connect enterprise data with AI capabilities in a secure, scalable manner. A hybrid working model is supported.

Qualifications

• Hands-on ETL/ELT pipeline development
• Data transformation, cleaning, modeling
• Strong SQL skills with large datasets
• Experience with Airflow, AWS Glue, S3/Redshift/Lambda
• Familiar with data quality, lineage, and reliability concepts

Responsibilities

• Own data lake and AI platform architecture and operations
• Build and maintain scalable data pipelines on AWS cloud
• Ensure data quality, governance, lineage, observability and security
• Establish standards for ingestion, transformation, storage, and consumption
• Develop AI agents, AI advisors and GenAI-powered solutions
• Design RAG architectures leveraging enterprise data assets
• Create multi-agent workflows to automate business processes
• Evaluate new AI tech and identify opportunities to enhance capabilities

Skills

Data Eng Fundamentals

Python/Backend

GenAI/LLM

Data Storage

DevOps/Platform

Tools

AWS Glue

Airflow

S3

Redshift

Lambda

Docker

Kubernetes

PostgreSQL

MySQL

Job description

UOB KAY HIAN PRIVATE LIMITED is seeking a hands-on data engineer to own the enterprise Data Lake and AI Platform. You will design and maintain scalable data pipelines using AWS Glue, S3, Redshift, and related cloud-native tools while ensuring data quality, governance, and security.

The role emphasizes building AI agents, GenAI-powered solutions, and RAG architectures to connect enterprise data with AI capabilities in a secure, scalable manner. A hybrid working model is supported.