Data Foundations Engineer – Scalable Pipelines & Open AI

Reflection; San Francisco (CA); On-site; USD 150,000 - 210,000; Full time

Benefits offered by this job

Top-tier compensation

Stock options

Health & wellness

Meals

22 weeks paid parental leave

Unlimited vacation (US)

Job summary

Reflection is seeking a data engineering leader to build core data systems powering research, training, and production environments in San Francisco. You will design reliable, scalable ingestion and orchestration for batch and streaming workloads, and develop storage and compute foundations for reproducible experimentation and rapid iteration.

The role emphasizes data quality, governance, and a unified data layer to support model development and production decision-making.

Qualifications

• Strong data engineering background with production-grade pipelines.
• Experience owning end-to-end data systems for large-scale workloads.
• Comfortable debugging pipeline failures and optimizing cost/performance.
• Bias toward action and impact in a fast-paced environment.
• Collaborative, clear communicator across research and infra boundaries.

Responsibilities

• Design ingestion and orchestration patterns for batch and streaming workloads.
• Build scalable compute and storage foundations for data processing.
• Ensure reproducible pipelines through versioning, backfills, and isolated execution environments.
• Provide trusted data quality, lineage, and governance signals for production decisions.
• Maintain a unified data layer across model development lifecycle.

Skills

Data engineering

Production pipelines

Cloud data platforms

Collaboration

Tools

Spark

Flink

Airflow

Dagster

Kafka

Pub/Sub

BigQuery

Snowflake

Parquet

Iceberg

Job description

Reflection is seeking a data engineering leader to build core data systems powering research, training, and production environments in San Francisco. You will design reliable, scalable ingestion and orchestration for batch and streaming workloads, and develop storage and compute foundations for reproducible experimentation and rapid iteration.

The role emphasizes data quality, governance, and a unified data layer to support model development and production decision-making.