Data & Analytics Lead, GreenNode

GreenNode is VNG's AI cloud business: public cloud & AI services business serving enterprise and AI-native customers across APAC. We are scaling from a regional cloud provider into an AI infrastructure company, and the data that runs the business has not kept up with the business. This role fixes that.

About the role

You will own GreenNode's business ontology: the entities (customer, contract and annex, product/SKU, tenant, capacity), the metrics and their definitions, the stewards accountable for each, and the semantic layer through which people, reports and AI agents query them. The charter is company-wide from day one. The hard part of this job is agreeing definitions across functions, not technology.

This is a business-first, hands-on role, not an engineering role. GreenNode already has data engineers and its own lakehouse product (Dataprise) in the Engineering team; you will use them. What we do not have is someone who can sit with Finance, RevOps, Sales Ops and Product, agree what the numbers mean, write that down as a model, and prove it on real data. You will do the definitions and the modelling yourself, and borrow engineering for pipelines and infrastructure.

What you will own

• Data inventory and demand map. A complete map of GreenNode's data sources (owner, grain, keys, refresh, quality), a map of consumers and the decisions they need data for, and a prioritised roadmap of data products.
• Business ontology and master data. The entity model and golden records for customer, contract/annex, product/SKU, tenant and capacity, with matching keys across the seven sources. One metric dictionary (ARR/MRR, bookings, churn/NRR, utilisation, revenue and cost by product/customer, cost per GPU-hour and per million tokens): each metric with definition, formula, source, dimensions and a named steward. Lightweight data contracts with producing teams (Sales Ops, billing, infra).
• Semantic layer. The governed layer that turns the ontology into something queryable: models, metrics and dimensions defined once and consumed by BI, management reporting, forecasting and AI agents (via MCP). You define it; engineering runs it.
• Delivery, with Engineering. The thin slices that bring sources onto the central store on Dataprise, GreenNode's own lakehouse platform, using the Engineering data pool. You write the acceptance criteria and the models; the pool builds connectors and infrastructure within an agreed capacity envelope.
• Governance. Stewardship, quality, lineage, access, retention and data-protection requirements, in partnership with GRC and the DPO

Yêu cầu

• 4-8 years in data, with at least one experience of defining and building a company's metric layer and entity model from scratch in a company of a few hundred people, and living with the consequences.
• Business-first: you define metrics with the CFO, sales ops and the product lead in the same week, you can explain why two teams' revenue numbers differ, and you can say no to a dashboard that has no decision behind it.
• Hands-on with data: fluent SQL and dbt-style data modelling. You write the model yourself and test your definitions on real data; you do not hand a document to engineering and wait.
• Experience integrating CRM + ERP + billing/usage data into one model; strong instincts for master data (matching keys, survivorship, stewardship).
• Comfortable operating through influence: a borrowed engineering pool, function heads who must agree definitions, producers who own their data quality.
• English working proficiency (written and spoken).

Nice to have

• Cloud / SaaS / infrastructure business metrics (ARR, NRR, utilisation, cost per unit); exposure to Zoho CRM, Oracle ERP/HFM extracts, or metering/billing systems.
• Semantic-layer and AI-consumption experience (dbt Semantic Layer, Cube, LookML, MCP servers, LLM-on-warehouse patterns); familiarity with ontology-style object models (Palantir Foundry or similar).
• Working knowledge of the platform side: ELT tooling, a cloud warehouse or lakehouse (Trino/Spark, Iceberg/Delta), BI tools, Python. Enough to write acceptance criteria for engineers, not to run the platform.
• Experience in a VNG-scale group environment (consolidation, entity structures, audit).

• A data-engineering or platform-engineering role: GreenNode's engineers build and run the pipelines and the platform.
• A BI/reporting factory, or a finance-only reporting role: the charter is company-wide.
• A documentation-only architect role: definitions that are not proven on real data do not count.
• A people-management-only role in year one.