First seen by Alion on Nov 21, 2026.
Responsibilities
About The Role
At Tentang Anak, decisions should start from data people trust, and that same data will increasingly power AI tools that help teams find answers faster. You’ll design and build the data warehouse that supports both: one that scales with the business, is governed from day one, and is secured down to the row. You’ll own the platform end to end, from the first source connection through to the semantic layer that analysts, dashboards and LLMs query.
Scope: you’ll sit within the Business Intelligence & Data Engineering team as an independent individual contributor, setting the data architecture and engineering standards the team builds on.
What You'll Own
- Build the warehouse foundation: design and implement a layered BigQuery warehouse (raw → staging → core → marts) with clear modeling standards and a cost-efficient structure.
- Make data AI-ready: build a documented semantic layer, consistent metric definitions and rich metadata, so LLMs and AI agents can query the data accurately and safely.
- Ship governance by design: set up data ownership, cataloging, lineage, classification of sensitive data, and access policies before the warehouse scales.
- Make pipelines reliable: build and orchestrate ELT pipelines with testing, monitoring and alerting, so failures are caught before stakeholders notice.
- Keep the numbers trustworthy: put automated data-quality checks, SLAs and freshness tracking on every critical dataset.
- Lock down access precisely: implement dataset-level, row-level and column-level security, so each team, partner or AI tool sees only the data it should.
- Control cost and performance: use partitioning, clustering and materialization strategies, and monitor query spend, so the platform scales without the bill scaling with it.
- Raise the bar: define engineering standards (version control, code review, CI/CD, documentation) and mentor analysts on writing production-grade SQL.
What Success Looks Like
- Core business domains run on the new warehouse, with documented models and a single source of truth for key metrics.
- A governance framework is live, covering data owners, a data catalog, lineage, and an access-request process.
- At least one AI use case, such as an LLM or agent answering business questions, runs on governed warehouse data.
- Critical pipelines meet agreed freshness SLAs and have automated quality tests and alerting.
- Dataset-level and row-level security are enforced on sensitive data and audited regularly.
Why Tentang Anak
Real ownership, a warehouse you get to build right from the start, close access to decisionmakers, and room to bring AI into how data gets done.
Requirements
Must-Have Requirements
- 3+ years in data engineering, including at least 1 year designing a data warehouse from the ground up rather than only maintaining one.
- Expert SQL and strong Python for building production pipelines.
- Experience building and running orchestrated pipelines in production, using Mage.ai or a similar tool (Airflow, Dagster, Prefect), with monitoring, testing and incident handling.
- Experience with a cloud data warehouse, preferably BigQuery, including partitioning, clustering and cost optimization.
- A proven track record of implementing data governance and access control, including row level and column level security in production.
- Strong dimensional modeling skills (Kimball, star schema, slowly changing dimensions) and experience with dbt or a similar transformation framework.
- Genuine curiosity about how AI and LLMs will consume data, and what that requires from a warehouse.
- The ability to explain architecture trade offs clearly to non-technical stakeholders.
Nice-to-Have
- Experience building semantic layers or metric stores (LookML, dbt Semantic Layer, Cube).
- Experience building AI assistants, agents, evaluations or model services.
- Experience preparing data for LLM or agent use, such as RAG pipelines, embeddings, vector search, or text-to-SQL evaluation.
- Experience in the FMCG or consumer industry.
- A GCP certification (Professional Data Engineer).

