**About the team**
Masan's centralized data platform powers analytics and machine learning across Vietnam's largest consumer-retail ecosystem (WinCommerce, WinX loyalty, Masan Consumer, Phuc Long and more). The data platform engineering team builds the layer that turns raw source data into contracted, monitored, reusable data products on Azure Databricks.
We are looking for a Data Engineer with 1–2+ years of experience to help build an agentic data platform: one that automates the process of building data products at scale, from ingestion and modeling to monitoring and ML forecasting. You will build the agentic workflows and data products that run on this platform, work alongside senior engineers who own its shared components, and use AI coding agents in your daily work. Our success is measured by how little of our code we write by hand.
**What you will do**
**Model and keep data trustworthy**
- Build dimensional models (grain, facts, dimensions, slowly changing dimensions) following the
team's dimensional-modeling procedure, starting with finance
- Build feature stores, feature engineering for ML models, operate BI reporting products across
business unit
- Add quality rules and SLAs to contracts with producer and consumer teams, and run the contract monitors
- Investigate monitor findings, backfill, and document confirmed insights in the business-unit knowledge base
- Define each KPI once in the metrics store, so dashboards, forecasts and AI agents read the same number
**Build and operate the agentic data platform**
The platform turns a data product spec into running pipelines. AI agents draft the ingestion, models, contracts and tests from the spec, deploy them as code on Databricks, and monitor them in production. Engineers review every step.
- Take a data product from spec to deployed jobs and schemas with the platform's agents and
CLI, dev before prod
- Review, test and correct what the agents generate before it merges, and report gaps in the agent skills
- Operate platform runs: watch deployments and agent outputs, triage failures, and keep CI green
- Maintance cost within the allocated budget, optimize pipeline and query performance, services
reliability
**Must-have**
- 1–2+ years in data engineering; personal projects count
- Spark hands-on, on-premise or on Databricks: DataFrame API, Spark SQL, partitions and joins, reading and writing Parquet, Delta or Iceberg
- Strong foundation in data engineering: batch versus incremental loads, idempotent jobs, schema evolution, data quality checks
- Strong foundation in database systems: SQL, indexes and partitioning, transactions, reading a query plan, normalized versus dimensional schemas
- Agentic automation workflow capability: has built or run a workflow with an AI coding agent
(Claude Code, Cursor, Copilot agent) or an agent framework, and can review and test what it produces
- Strong problem-solving and critical-thinking skills: breaks down unclear problems, questions assumptions, and checks results against the data
**Priority given to candidates with**
- SAP ECC or S/4HANA ingestion experience
- Data migration experience: moving tables or pipelines from one platform to another with reconciliation of the result
**Nice to have**
- Databricks specifics: Unity Catalog, Jobs, Azure ADLS Gen2
- Open Data Product Specification (ODPS) or Open Data Contract Standard (ODCS)
- Infrastructure as code for Databricks: Pulumi, Terraform or Asset Bundles, deployed through
CI/CD pipelines such as GitHub Actions
- Streaming pipelines and online feature serving (a big advantage): building real-time pipelines with Structured Streaming, Auto Loader, Kafka or Lakeflow Declarative Pipelines, and serving fresh features to ML models at low latency
- Personal-data handling: masking, tagging and access rules