Data Platform Engineer

The role

This engineer joins the data engineering group to lay groundwork for Figma's AI-driven products and internal data experiences. The position is an early-stage seat shaping the company's machine learning and data platform. Work ranges from self-serve analytics to production-grade model systems.

What you would do

• Drive the AI data agent that lets people query data on their own, covering the agent layer, prompt pipelines, and usage tracking.
• Run the ML platform end to end: model serving, feature flows, job orchestration, deployment pipelines, and production health checks.
• Create data-facing products so models and datasets become core parts of the user experience.
• Build tooling that lets data scientists ship, adjust, and maintain models, including feature stores and staged rollouts.
• Architect infrastructure for conversational and AI-assisted querying across the organization.
• Coordinate platform work across data science, ML, infrastructure, and product groups, agreeing on data contracts and service levels.
• Make the platform easier for ML and data practitioners to use through reusable abstractions and capabilities.

What they asked for

• Five or more years in data platform, infrastructure, or ML engineering, including at least one year on AI or ML systems.
• Hands-on experience running the full ML lifecycle in production: training, evaluation, deployment, and monitoring.
• Solid Python skills, or a comparable language, with a record of dependable, scalable services.
• Background designing ML infrastructure such as serving layers, feature pipelines, and orchestration, plus scalable architectures.
• Track record of delivering projects with data science, engineering, infrastructure, and product teams.
• Skill in data modeling and shaping data-driven products.

Nice to have

• Familiarity with ML platform tools such as MLflow, Kubeflow, feature stores, or ML CI/CD.
• Exposure to LLMs, retrieval-augmented generation, prompt processing, or AI-native infrastructure.
• Prior work building self-serve analytics or internal data tooling.
• Working knowledge of current data tooling like Snowflake, dbt, Dagster, or AWS.
• A product-oriented outlook that ties platform work to user and business outcomes.