Staff Analytics Engineer — Data Warehouse

About the Role Together AI is building high-performance AI inference infrastructure and the software platform around it. We're looking for a senior Analytics Engineer who sits at the intersection of data engineering and business intelligence — someone who can turn raw, complex data into clean, trusted, well-documented models that the whole company can reason from. You'll own the transformation layer of our data warehouse: shaping bronze/silver/gold models, designing dimensional schemas, and acting as the connective tissue between engineering systems and business stakeholders. You are equally comfortable deep in a dbt project and in a room with Finance, GTM, and Product aligning on definitions. Requirements Expert SQL : window functions, complex aggregations, query optimization, cost-aware pattern selection, proficiency in Snowflake or equivalent cloud warehouse. dbt : deep, production-grade experience — models, tests (singular + generic), docs, snapshots, macros, packages, and incremental strategies. You've designed a dbt project from scratch and maintained it in production. Airflow / Astronomer : production DAG authoring, backfill handling, reliability patterns, and the Cosmos dbt integration. Dimensional modeling : you've read Kimball (or absorbed the equivalent), know the difference between star and snowflake schemas by feel, understand slowly changing dimensions, and can explain why a fact table's grain matters. Stakeholder management : demonstrated experience partnering with non-technical stakeholders, driving metric alignment, and delivering trusted data products — not just pipelines. Strong written communication: your documentation and async updates are clear enough that people don't need to ask follow-up questions. Strong plus Experience with financial data or billing data — ARR, usage-based billing, invoice reconciliation, revenue recognition patterns. We operate a usage-based inference billing system and this context transfers directly. Experience with PII handling, data masking, access-tier modeling, or compliance work (SOC 2, ISO 27001, GDPR, CCPA). Familiarity with lakehouse patterns (Iceberg, Delta, Hudi) and hybrid warehouse/lake architectures. Python for data tooling: automation, data quality frameworks, custom dbt macros or operators. Experience with Hex, Metabase, or similar notebook/BI tooling that sits on top of your dbt models. Prior experience in a high-growth AI/ML infrastructure or platform company. Responsibilities Modeling transformation Own and evolve the dbt transformation layer: design, implement, test, document, and maintain modular dbt projects that cover billing, product usage, financial data, and operational metrics. Build analytics-ready dimensional models following Kimball methodology: star schemas, conformed dimensions, fact tables with the right grain, and SCD Type 2 for slowly changing entities. Design for correctness, performance, and cost — partition strategies, incremental models, and avoiding full-table scans. Build and maintain a semantic/metrics layer with consistent, auditable metric definitions reused across notebooks, BI, and APIs. Pipeline orchestration Author and maintain Airflow DAGs (Astronomer-managed) that orchestrate dbt runs, data quality checks, and downstream dependencies reliably. Apply solid DAG design: idempotent tasks, proper backfill strategies, SLA alerting, and clean dependency graphs. Work in our Cosmos (dbt + Airflow) integration — you know when to use a DbtTaskGroup vs a custom operator. Data quality governance Implement data quality checks at every layer: freshness, null/uniqueness tests, referential integrity, distribution drift, and business-rule assertions. Drive data stewardship practices: ownership, SLAs, clear "source of truth" definitions, and change communication. Handle PII fields correctly — masking, anonymization, and access-tier alignment. Compliance experience (SOC 2, ISO 27001, or similar) is a meaningful plus. Sta ... (truncated, view full listing at source)