Senior Data Engineer - Fulltime - Worldwide

About the company

We're hiring two people into this role. Same scope, same bar; two seats on the data platform.

Roam builds the operating system for real estate organisations: one place where a enterprises own data, the market's data, and AI agents come together. Loci is Roam's market world model: continuously ingested and structured real estate data covering properties, listings, prices and rents, regions, regulation, demographics, and planning, exposed through tools, APIs, and agents.

We are hiring two senior data engineers to accelerate Loci under our data Lead. The work is data engineering: get data science out of notebooks, schedule and operate pipelines, and expose a stable contract so apps and agents can rely on the world model.

What you'll do

• Production MLOps for Loci: CI/CD for data and ML pipelines, Airflow orchestration and scheduling, reliable scheduled runs instead of ad hoc notebook jobs
• Take data science models and pipelines from research into production — versioned, tested, monitored, and runnable by the team without the original author
• Build and harden the Loci Data API and MCP tool contracts so Roam OS apps, visualisations, and agents can call market data with typed, versioned schemas, auth, and clear rate limits
• Advance the Loci lake and pipeline backlog: PySpark / Delta work (including market snapshot migration), entity resolution and deduplication, SQL modelling, and data-quality checks and alerts
• Harden Loci service access for client deployments: authentication and tenant context, controlled tool exposure, endpoint protection
• Ship through GitHub PRs reviewed by Roam engineers; leave documentation, contract notes, and short walkthroughs so the team can operate and extend what you build

Must-haves

• Senior production data engineering — pipelines and models from prototype to scheduled, monitored production
• Airflow (or equivalent, ready to work in Airflow day one)
• CI/CD for data and ML (MLOps): testing, packaging, promotion, rollback
• Building APIs and MCP servers that expose data and models (Python, FastAPI or similar; typed schemas, versioning, auth)
• Strong Python and SQL; solid judgement on data modelling and data quality
• Nice to have: Kubernetes, Databricks, Spark at scale / PySpark and Delta Lake, PostGIS or real estate data, Supabase or Postgres RLS, LLM evaluation or agent-facing data contracts