About the RoleYou own LILT's data transformation layer: the dbt layer and warehouse behind every number LILT reports, from Analytics to our LLM/MCP surface to internal dashboards. Metric definitions are often owned by other teams; you implement and keep them consistent. This layer has no owner today; you make it a role.
You lead a new Data sub-team in Platform Engineering, reporting to the head of Platform, as a hands-on player-manager while hiring and growing a Data Engineer and Senior Data Scientist. You hold decision rights over the transformation layer and warehouse and own their cost.
What You're Walking IntoWe want to be direct about this role so the right person applies.
- You inherit ambiguity. No single owner, pipelines to document and rebuild, and no team until your first two hires; until then you write the SQL, dbt, and Python yourself.
- You settle the numbers. Finance, Operations, Production, and Product must trust the same metrics; you keep definitions consistent and say no when needed.
- Some things are fixed, most aren't. dbt, a single warehouse, and on-prem parity are non-negotiable; warehouse cost is measured and expected to go down. Everything else is yours to decide, with a written case.
The Stack- Transformation: dbt on BigQuery
- Analytics serving: ClickHouse, Cloud for SaaS, self-hosted on-prem
- Sources: MySQL, replicated to BigQuery
- ETL/orchestration: Python 3, Argo Workflows on Kubernetes
- Consumers: In-app Analytics, Sigma, LILT's Assist agent, LILT's MCP server
- Observability: Datadog
- Agentic engineering: Claude Code and Cursor, used daily across Engineering
Key Responsibilities- Own the data layer. Implement every metric definition once in dbt, consistent everywhere it's used, partnering with the teams that define them. Business Operations, Production, Finance, and Product get one point of accountability; discrepancies resolve at the definition.
- Run the transformation layer and warehouse: every pipeline has an owner, tests, and a known cost; spend is measured and goes down.
- Set direction: warehouse strategy, ClickHouse's role, and how the layer is exposed via API and MCP, each backed by a written case.
- Build the team: hire a Data Engineer and a Senior Data Scientist, set the charter, and run delivery, quality, and on-call health.
- Set agentic engineering practice: define how the Data team uses AI agents to build, test, and review pipelines and models, including where human review is required.
- Stay hands-on: read, review, and write the SQL, dbt, and Python your team ships.
Qualifications- People management: 7+ years in data/analytics engineering, including 2+ years managing a small team (2-5), with a track record of hiring and developing ICs.
- Hands-on fundamentals: fluent in SQL and Python; has built and run production pipelines and a dbt (or equivalent) transformation layer; comfortable with BigQuery, ClickHouse, Snowflake, or similar.
- Cost and roadmap ownership: has owned a warehouse or pipeline budget and reduced it with measurable results; translates business needs into a technical plan and sequences a backlog against limited headcount.
- Stakeholder and business metrics: has owned data accountability for finance, operations, and go-to-market stakeholders, and understands B2B SaaS metrics (ARR, ACV, gross margin, on-time delivery) and how definition drift breaks them.
- Effective AI use and communication: uses AI tools daily and knows where they help, mislead, and need verification; documents decisions clearly and communicates tradeoffs, risk, and cost crisply to leadership.
Preferred Skills- Stood up a data function from zero, or revived an abandoned one.
- Run dbt in production at scale on BigQuery; operated ClickHouse.
- Shipped analytics that runs in both cloud and self-hosted environments.
