Principal Data Engineer

Hims & Hers is a health and wellness platform providing personalized access to healthcare, prescription, and wellness products. The Principal Data Engineer will set the technical direction for the company’s data platform, defining architecture, streaming strategy, governance, reliability, economics, and self-service capabilities. The role also provides technical leadership through architecture reviews, documentation, mentorship, production coding, and systemic incident resolution.

Responsibilities:

• The multi-year platform architecture. Ingestion, streaming, orchestration, transformation, and serving as one system rather than five. You decide what we build, what we buy, what we deprecate, and in what order
• The streaming bet. Whether Kafka to Flink to BigQuery stays an analytics pipeline or becomes production infrastructure, and what that commits us to operationally, financially, and organizationally. You are the architect of the answer you choose
• The contracts between organizations. The interfaces, service boundaries, and data contracts govern how other teams consume from and write into the platform. Where your standards end and another org's autonomy begins is a boundary you define with the VP of Data, not one handed to you
• What "trustworthy data" means here, and who is on the hook for it. You define the quality and freshness guarantees the platform makes, the tiering that determines which datasets get them, and the accountability model when they're missed. Staff engineers build the detection, validation, and alerting systems. You decide what we promise, to whom, and at what cost
• The economics of the platform. The design decisions that determine what data costs us at scale (storage and compute strategy, reservation model, orchestration footprint) and the accountability framework that keeps consumption honest as more teams build on it. You set the framework. You don't tune every query
• The self service strategy. What capabilities let Analytics Engineering and Data Science move without waiting on DPE, what guardrails keep that from breaking production, and how the platform stops being a bottleneck as demand grows faster than headcount
• The standards the discipline runs on. Testing, CI/CD, observability, schema governance, and infrastructure as code practice, adopted by teams that don't report to you because the standards are good rather than because you can compel them
• Architecture Review. You chair it, and you are the decision-maker of record for cross-team, multi-system, and cost-impacting changes, including ones that touch systems you don't own
• The seams between data and everything else. ML and Data Science readiness, HIPAA and PHI compliance posture with Legal and Security, and infrastructure hardening with DevOps. Nobody else is looking at all four at once
• The written record. ADRs, design docs, and RFCs that other teams cite as the reference. If a decision only exists in your head, you haven't finished making it
• The technical ceiling of the org. Mentorship, design review, and hands-on pairing that makes Staff and Senior engineers better. Part of how you'll measure your impact is the volume of decisions you no longer have to make yourself
• You stay hands-on throughout. You write and review production code, and you're in the room for platform-level P1s, driving the systemic fix rather than owning the runbook. This is a technical authority role, not an advisory one

Requirements:

• 12+ years building and owning data platform architecture, with a record of decisions whose consequences you stayed to live with, ideally across multiple companies or multiple platform generations
• Evidence you've set technical direction for organizations you didn't manage, and got adoption through credibility rather than reporting lines
• Experience defining the edges of your own architectural authority. Ambiguity about where your mandate ends should not stall you
• Deep GCP and BigQuery expertise, plus real operational fluency in AWS (our Airflow runs on EKS). Multicloud is the job here, not a bonus
• Modern data stack at platform scale: governing dbt across teams, Airflow/Astronomer, Kafka/Confluent, Databricks/Spark, Fivetran, and reverse ETL or activation platforms such as Hightouch
• Event streaming in production at scale, including Schema Registry, data contracts, consumer lag management, and delivery semantics along with what they actually cost. Bonus if you've personally made the analytics versus production streaming call
• Data governance and compliance in a regulated environment: HIPAA and PHI handling, data classification, access controls, audit logging, and GDPR
• Terraform or equivalent, and the conviction that infrastructure changes are software changes
• Strong Python and SQL. You raise the bar on production pipeline code by reviewing it, not by talking about it
• Writing that creates alignment across organizations. Your design docs are where decisions get made, not summaries of decisions already made elsewhere
• Databricks, Unity Catalog, and Delta Lake in production at scale
• CDC patterns and Flink for real-time processing, particularly where streaming infrastructure served both analytics and production or model serving paths across multiple teams
• PySpark and SparkSQL for large-scale batch and streaming workloads
• A BigQuery to Databricks Lakehouse migration, or an equivalent warehouse migration you led end-to-end
• MLOps partnership: model training pipelines, feature stores, experimentation infrastructure, or the data foundations behind AI-driven product features
• Go or Python service development for Kafka producers and consumers
• Direct-to-consumer healthcare or telehealth with HIPAA and GDPR obligations
• SOX compliance controls in a data engineering context