AI Native Engineer

Unilabs seeks a Senior AI Native Engineer in Zürich to build agentic AI systems for clinical data extraction and diagnostic pipelines, combining entrepreneurial ownership with large-scale healthcare data access.

This position focuses on developing agentic AI solutions that directly influence precision clinical trials, therapeutic development, and preventative health markets. The role provides significant ownership, allowing the engineer to drive execution and impact within a rapidly scaling environment backed by a major global diagnostics group.

Responsibilities

• Create and sustain production-ready LLM pipelines designed to automatically process large volumes of unstructured PDF pathology reports.
• Extract key clinical information, including diagnoses, tumor grades, staging details, and biomarker statuses, from raw text documents.
• Assess and deploy specialized agentic frameworks or orchestration tools, ensuring extraction accuracy meets clinical standards.
• Develop confidence scoring mechanisms and human-in-the-loop review queues to handle low-confidence extractions based on clinical parameters.
• Design data pipelines that integrate pathology laboratory information system (LIS) data with molecular and genetics databases.
• Implement secure interfaces using REST APIs, HL7 v2, or HL7 FHIR to connect structured data with downstream ecosystems like Proscia Concentriq and Aperture.
• Establish the architectural foundation for high-throughput API layers that will connect with consumer wearables, health apps, and hospital systems.
• Build automated monitoring systems to detect data quality issues, missing fields, or incomplete records before they reach delivery endpoints.
• Apply technical de-identification methods to remove or pseudonymize patient identifiers in compliance with privacy regulations.
• Ensure all technical work adheres to strict health data privacy standards, including Swiss nDSG and EU GDPR Article 9.
• Maintain comprehensive audit logs and data lineage tracking to preserve clinical data provenance for partners and pharma clients.

Requirements

• Hands-on experience using LLM APIs, designing system prompt state machines, and fine-tuning prompts for structured text extraction.
• Proven ability to work with agentic frameworks like LangChain or LlamaIndex to orchestrate complex, multi-step clinical data workflows.
• Experience deploying non-deterministic models in production, including managing context windows, token costs, rate limits, and evaluation metrics.
• 4–7+ years of software engineering experience with strong proficiency in Python and SQL, capable of independently debugging asynchronous pipelines.
• Deep knowledge of building and consuming production-grade REST APIs within regulated or clinical settings.
• Practical experience deploying applications on cloud platforms (AWS, Azure, GCP) using Docker containerization.

Nice to have

• Familiarity with clinical health standards such as HL7 v2, FHIR, OMOP CDM, or CDISC conventions.
• Experience with digital pathology data formats (DICOM, SVS, NDPI) or laboratory information systems (LIS).

What the company offers

• Hybrid work model offering a mix of office and remote flexibility.
• International, collaborative, and highly regulated product environment.
• Competitive compensation and benefits package.
• Long-term ownership of a strategic healthcare product with equity upside.
• Direct access to Europe's largest diagnostic data pool, including pathology, imaging, and blood test data from millions of patient journeys.
• Opportunity to deploy agentic AI that impacts precision clinical trials, drug development, and preventative longevity markets.
• Entrepreneurial ownership and execution speed similar to a seed-stage startup, supported by the infrastructure of Unilabs and A.P. Møller Holding.

About the company

Unilabs is one of Europe's leading diagnostics groups, operating over 200 laboratories across 14 countries and performing more than 237 million diagnostic tests annually. The company fosters a modern engineering culture that values shipping impactful products over manual effort, encouraging the use of AI-assisted tools to accelerate problem-solving.

• Emphasis on shipping impactful products rather than manual effort.
• Encouragement to use AI-assisted development environments like Cursor or GitHub Copilot to boost productivity.
• Modern tooling and collaborative international work environment.