Data Engineer

📋 Description

• • Architect and own end-to-end data pipelines that ingest, cleanse, and transform multi-terabyte healthcare datasets (claims, eligibility, Rx, lab, provider directory, and patient-generated data) into Nomi’s cloud-native platform, ensuring 99.9 % uptime and sub-5-minute freshness for critical analytics.
• • Build scalable ETL/ELT workflows in Python and SQL (Snowflake) that automate complex business rules, HIPAA-compliant de-identification, and real-time anomaly detection; leverage AWS Lambda, Step Functions, and EventBridge to orchestrate serverless jobs that cut processing costs by 30 % while doubling throughput.
• • Partner with product managers, actuaries, and clinicians to translate vague requirements into concrete data models—designing star and snowflake schemas, slowly-changing dimensions, and NoSQL (DocumentDB) collections that power executive dashboards, provider scorecards, and predictive-risk algorithms used by 30 M+ members.
• • Implement data-quality frameworks (Great Expectations, dbt tests) that surface schema drift, null-rate spikes, and payer-specific formatting errors within minutes; maintain SLAs that keep duplicate rate <0.1 % and data lineage searchable down to the column level for auditors and downstream ML engineers.
• • Champion a “data-first” culture: conduct brown-bag sessions on GitFlow, Jira story slicing, and peer-review best practices; mentor junior engineers on writing idempotent transforms, unit-testing SQL, and choosing the right storage layer (S3 vs. Redshift vs. DocumentDB) for each access pattern.
• • Own the DevOps lifecycle—spin up infrastructure-as-code (Terraform) for isolated dev/stage/prod environments, configure CI/CD via GitHub Actions, and monitor pipelines with CloudWatch, Datadog, and PagerDuty to achieve <30-minute MTTR when alerts fire at 2 a.m.
• • Collaborate with security and compliance teams to encrypt PHI in transit and at rest, apply role-based access controls down to the row level, and generate audit trails that satisfy SOC-2 and HITRUST evidence requirements without slowing down agile releases.
• • Experiment relentlessly: prototype Spark on EMR for large-batch retro-processing, evaluate dbt vs. Dataform for declarative modeling, and pilot Kafka/Kinesis streaming when batch windows can’t keep pace with same-day employer reporting demands.
• • Influence the roadmap—surface pain points like payer API throttling or stale provider rosters, quantify the ROI of faster data, and pitch epic-level stories that secure engineering sprint capacity and cloud budget for the next quarter.
• • Celebrate wins together—whether it’s a 50 % reduction in claim-adjudication latency or the launch of a public-price transparency tool that lands Nomi on the front page of Healthcare IT News, you’ll share the spotlight with a mission-driven team that believes healthcare can be fixed.