Senior AI Analytics Engineer
- Hybrid
- Hong Kong
- senior
Posted last checked
Key Responsibilities - Data & AI Architecture: Design, build, and maintain production-grade data models to bridge raw data streams with downstream applications and BI tools. Using LLMs and AI tools to generate data documentation, write SQL/Python transformations, convert warehouse dialects, and diagnose pipeline error logs.
- Analytics Engineering: Standardize data modeling practices and build robust, reliable, scalable semantic layers that feed clean, trustworthy data to downstream analytics and reporting.
- Business & Requirements Analysis: Occasionally lead scoping discussions with product, tech and finance stakeholders to translate business requirements into functional specs, user stories, and data mapping documents.
- Business Logic Translation: Reconciling data definitions across departments, a critical human-led task that AI cannot fully automate
- Data Quality & Governance: Establish automated testing, monitoring, and data quality frameworks to guarantee financial-grade reliability.
- Performance Optimization: Optimize query performance, warehouse costs, and pipeline orchestration.
Requirements & Qualifications
- Experience: 5+ years of experience in Data Engineering, Analytics Engineering, or Data Science, with at least 2 years in an AI environment.
- Technical Mastery:
Advanced SQL capabilities.
Hands-on expertise with modern data transformation and orchestration frameworks (dbt, Airflow, Prefect, Dagster).
Deep experience with cloud data warehouses (Snowflake, Databricks, BigQuery).
- AI Technical Stack:
Understanding of feature engineering, AI agentic workflow lifecycles, and evaluation metrics.
Experience with deploying AI agentic workflows in production environments.
- Business Analysis Skills: Demonstrated ability to occasionally gather business requirements, define key metrics, and translate stakeholder needs into structured data assets.
- Software Engineering Principles: Strong experience with version control (Git), CI/CD practices for data pipelines, and production code testing.
