Senior Data Engineer

Tempus AI•Chicago, IL

About The Position

Passionate about precision medicine and advancing the healthcare industry? Recent advancements in underlying technology have finally made it possible for AI to impact clinical care in a meaningful way. Tempus' proprietary platform connects an entire ecosystem of real-world evidence to deliver real-time, actionable insights to physicians, providing critical information about the right treatments for the right patients, at the right time. We are building the Patient Evaluation Engine: a high-scale, multi-modal healthcare platform where autonomous AI agents reason over clinical data to drive real-time clinical evaluation across federated networks of hospitals. We are looking for a Senior Data Engineer to build and own the data platform underneath it — the pipelines, models, and services that make EHR records, genomic results, and cardiovascular imaging discoverable, trustworthy, and usable by agents. This is a data engineering role at its core, and it asks for two things beyond the usual scope. First, you should be a capable software engineer: the person who builds the pipeline here is the person who writes the service that exposes it, and you will regularly work in our TypeScript application and service code rather than handing that off. Second, you should know cloud infrastructure well, specifically Google Cloud — you will make real decisions about how this platform is deployed, scaled, secured, and paid for, not just what runs on it. Our goal is to move beyond static data warehousing toward a dynamic, "agent-ready" data fabric that supports real-time clinical evaluation at enterprise scale, in a HIPAA-regulated environment. The platform is early and much of it is still being built, which is why we are looking for someone with high ownership and a strong self-starting instinct rather than someone waiting for a fully specified backlog.

Requirements

Data engineering depth. Proven track record building and operating production data pipelines that handle structured and unstructured data at scale, with real ownership of reliability and correctness.
Google Cloud fluency. Hands-on experience designing and running workloads on GCP — BigQuery, Pub/Sub, Cloud Storage, Cloud SQL, and Secret Manager — including the IAM and service-account model that controls access to sensitive data.
Analytics engineering. Strong SQL and practical experience with dbt or an equivalent transformation framework, including testing, documentation, and managing a model layer as it grows.
Infrastructure practice. Comfort owning infrastructure as code in Terraform, working in containers, and taking responsibility for the operational characteristics of what you deploy.
Software engineering ability. You write production-quality application and service code, not just pipeline glue — including APIs, tests, and the design work that goes with them.
Python and TypeScript. Python strong enough for production pipelines as well as hands-on data profiling and debugging, plus enough TypeScript or another statically typed language to work confidently in our service and application code.
Event-driven systems. Experience with pub/sub or queue-based architectures and the failure modes that come with them — retries, ordering, idempotency, and dead-letter handling.
Interoperability standards. Working knowledge of HL7, FHIR, and Epic/Cerner data structures, along with DICOM and genomic data formats.
Regulatory fluency. Familiarity with building secure, resilient systems under HIPAA and SOC 2.
Total Professional Experience: 5+ years building data-intensive software systems in production.
Data Engineering: 3+ years focused on data engineering, pipeline ownership, or data modeling, ideally in the healthcare or life sciences domain.
Cloud Infrastructure: 2+ years hands-on building and operating on Google Cloud, with demonstrated ownership of infrastructure decisions rather than consuming someone else's.
Healthcare Domain: 2+ years in HIPAA-regulated environments, with hands-on exposure to EMR integrations (Epic, Cerner) and healthcare data standards.
AI/ML Orchestration: 1+ years hands-on building with Large Language Models — agentic workflows, RAG, or autonomous tool use.
Data at Scale: Demonstrated experience managing structured (SQL, NoSQL) and unstructured data at a scale of millions of records, ensuring data integrity for downstream AI consumption.

Nice To Haves

Google Cloud Healthcare API. Direct experience with managed FHIR or DICOM stores.
Specialized clinical data. Direct experience with OMOP, DICOM, genomic data models, or longitudinal patient records.
Kubernetes. Experience running containerized workloads on Kubernetes.
AI data engineering. Experience with vector databases (Pinecone, Weaviate, pgvector) or graph databases to support RAG and agentic memory.
AWS. Experience with AWS services alongside GCP in a multi-cloud environment.
Advanced modeling techniques. Experience with Data Vault 2.0, Master Data Management, or comparable enterprise modeling methodologies.

Responsibilities

Build the pipelines that feed the agents.
Develop the systems that fetch, parse, and serve both structured and unstructured data in formats optimized for real-time inference — spanning clinical EHR records, high-throughput genomics (NGS), and cardiovascular imaging (Echo, Cath, ECG).
Own the warehouse and its transformations. Build and maintain our dbt models on BigQuery, along with the tests, documentation, and SQL standards that keep a growing model layer trustworthy.
Model the multi-modal patient record. Shape the data model across those domains, applying normalized and dimensional design as each one demands, and write the code that enforces it.
Move data through event-driven services. Build and operate the Pub/Sub topics, subscriptions, and dead-letter handling that connect ingestion, evaluation, and result delivery, with the retry and idempotency behavior that reliability at scale requires.
Make the data agent-ready. Build the data access patterns and metadata layers that let AI agents autonomously discover, query, and reason over structured and unstructured datasets, and the retrieval services those agents call.
Write the software on top. Build the TypeScript services and APIs that handle agent input and output and coordinate specialized agents, meeting the platform's performance and scalability demands. You are expected to be comfortable in the application codebase, not only in the data layer.
Own the infrastructure your platform runs on. Extend and operate the platform's Google Cloud footprint in Terraform — BigQuery datasets, Pub/Sub, Cloud SQL, Cloud Storage, Memorystore, Secret Manager, and the service-account and IAM model that governs access to clinical data.
Scale across hospital networks. Build for federated networks of hospitals: multi-tenancy, high availability, and performance across hybrid on-prem and cloud environments built for sensitive health-system integrations.
Guarantee ground truth. Implement automated solutions to monitor data quality and lineage with strict traceability back to source systems, ensuring "ground truth" for agentic evaluations.
Instrument for trust. Build the observability, error tracking, and human-in-the-loop checkpoints that make automated clinical evaluation transparent and debuggable.
Raise the standard around you. Partner with clinical, analytics, and platform engineering teams on data modeling standards, governance, and practices for maintaining data integrity in a HIPAA-regulated environment.