Data Engineering Tech Lead

Expion Health is building a pharmacy economics platform that connects pharmacy value, cost, clinical decisions, and performance. The Data Engineering Tech Lead will own the technical direction, quality, and delivery of a data engineering pod supporting claim repricing, savings analytics, client reporting, and AI/ML models. The role involves designing Snowflake and AWS data platforms, building healthcare data pipelines, enforcing data quality and compliance, mentoring engineers, and partnering with analytics and business stakeholders.

Responsibilities:

• Design and evolve the data platform on Snowflake and AWS, including ingestion, storage, orchestrated transformation, and the serving layer feeding ExpionIQ analytics, client reporting, and ML/AI models
• Own Snowflake architecture end to end: database and schema design, Snowpipe ingestion, in-warehouse transformation with Snowpark and dbt, RBAC and PHI-safe access patterns, and warehouse sizing and credit consumption against a budget
• Make and document the build-vs-buy and pattern decisions across batch versus streaming, ELT in Snowflake versus external Spark, modeling, and CDC, and drive them to consensus with architecture and security
• Modernize legacy claim-processing data flows onto repeatable, testable pipelines without disrupting production repricing volume
• Ingest and normalize high-volume healthcare data, including X12 EDI (837 claims, 835 remittance, 834 eligibility), provider and facility rosters, CMS and state fee schedules, contract and network terms, NDC and drug pricing files, and rebate and invoice data
• Build dimensional and semantic models that let analysts answer how much was saved, on which claims, versus what benchmark, without reverse-engineering SQL
• Support AI and automation workflows such as document understanding, claim reconciliation, and anomaly detection with clean, well-labeled, reproducible training and inference data
• Stand up data quality as a first-class system: contracts, expectations, reconciliation controls against source-of-truth totals, freshness and volume SLAs, and alerting that pages a human before a client notices
• Own lineage, cataloging, and documentation so pricing logic is auditable end to end
• Enforce PHI and PII handling by design under HIPAA, including least-privilege access, encryption, tokenization or de-identification, and retention rules
• Mentor junior and mid-level engineers day to day through pairing, teaching-oriented code review, and direct, useful feedback
• Assign work deliberately for development, stretching engineers onto designs and pipelines you could have written yourself while staying available to unblock them
• Onboard new engineers into the claim and pharmacy data domains and build the documentation, runbooks, and reference implementations that shorten the ramp for the next hire
• Set and enforce engineering standards for code review, testing, CI/CD, infrastructure as code, and observability, and reduce key-person risk across the pod
• Run technical design reviews, break roadmap epics into estimable work, and serve as the escalation point for production data incidents
• Partner with analytics and BI teams to co-design the semantic and reporting layer and agree on metric definitions so savings means the same thing in every dashboard, client report, and model
• Sit with pharmacy and medical cost-management, operations, and client-facing teams to understand the workflows behind the data before designing for it
• Translate in both directions, turning ambiguous business asks into concrete data requirements and explaining technical trade-offs in terms stakeholders can decide on
• Support client onboarding and audit or reporting requests as a partner to the business, turning recurring one-off asks into self-service data products
• Take on other work as needed to support the broader goals of the department and the company
• Standardize ingestion and transformation patterns and stand up automated data quality checks and SLAs on the highest-risk pipelines
• Reduce data incidents and manual reconciliation while modernizing legacy claim-processing flows without disrupting production volume
• Build a scalable, documented platform that supports new client onboarding and new savings products with materially less engineering effort per client

Requirements:

• 8 or more years in data engineering, including 2 or more years as a tech lead, staff engineer, or equivalent technical owner of a team's delivery
• Expert SQL and strong Python, with production experience building and operating distributed data processing such as Spark or AWS Glue
• Deep hands-on Snowflake experience at scale, including warehouse sizing and cost control, RBAC, secure data sharing, streams and tasks, time travel, query tuning on large claim tables, and clustering strategy
• Snowpipe experience, including continuous and auto-ingest loading from S3, Snowpipe Streaming, error handling and reconciliation, and knowing when Snowpipe fits versus batch COPY or external tables
• Snowpark in Python for pushing transformation and feature engineering into Snowflake, including UDFs, UDTFs, and stored procedures
• Deep AWS data stack experience across S3, Glue, Lambda, Step Functions, and RDS, with IAM and networking fundamentals and AWS-to-Snowflake integration patterns
• Proven dimensional and warehouse modeling on messy, high-cardinality real-world data, with the ability to defend your grain choices
• Production orchestration and transformation tooling such as Airflow or Step Functions and dbt on Snowflake
• Software engineering discipline applied to data: version control, automated testing, CI/CD, infrastructure as code, and monitoring and alerting
• Demonstrated ownership of data quality and observability in a system where wrong numbers have external consequences
• A track record of mentoring junior engineers, with specific examples of people who got measurably better working with you
• Proven ability to work directly with analytics and non-technical business stakeholders, gathering requirements, aligning on definitions, and managing expectations
• Clear written communication across design docs, runbooks, and incident write-ups that people actually use
• Healthcare payor or RCM data experience, including claims (837/835), eligibility (834), CPT/HCPCS/ICD-10/DRG/NDC coding, provider data, fee schedules, and reference-based pricing
• PBM or pharmacy data experience, including formulary, rebates, specialty drug pricing, and claim-level adjudication data
• Working knowledge of HIPAA in practice, plus SOC 2 or HITRUST control environments
• Snowflake Cortex, Snowflake ML, or Snowpark ML for in-warehouse AI/ML workloads, plus streaming patterns such as Kinesis or Snowpipe for near-real-time delivery
• Experience migrating legacy ETL or another warehouse onto Snowflake, and a Snowflake cost optimization track record