AI Bioinformatics Benchmarking Engineer

The Position

At Sequencing, you'll build and operate the evaluation and validation infrastructure that keeps our AI systems correct, reliable, and regression-protected. This is an execution-focused role: you'll design, implement, and maintain the benchmarking systems that continuously test AI outputs against curated genomic datasets, partnering closely with the AI Bioinformatics Engineering Lead to turn interpretation standards into measurable, operational systems.

The Impact

• Build and run automated evaluation pipelines for AI outputs end-to-end.
• Expand regression datasets that cover variant normalization, transcript ambiguity, and disease-level mapping.
• Execute large-scale validation runs across curated VCF datasets, track performance over time, and surface accuracy regressions the moment they appear.
• Grow a structured question bank using DV3 and related datasets, encoding the edge cases that matter most.
• Maintain ground truth datasets for pathogenic and benign variants so our benchmarks reflect real-world genomic complexity.
• Ship automated regression tests for every AI release and build dashboards that make variant-level and case-level accuracy visible to the whole team.
• Partner with engineering to wire validation directly into CI/CD, and diagnose and document failure modes as they emerge.

Dominant and Expressed Traits

• A degree in Bioinformatics, Computational Biology, Genetics, or a related field.
• Deep, hands-on experience with VCFs and real-world genomic datasets.
• Fluency with ClinVar, dbSNP, HGVS standards, and transcript mapping.
• A track record building evaluation frameworks or testing pipelines, Langfuse or similar.
• Experience evaluating LLM-based systems and standing up automated QA at scale.
• Comfort living inside modern data engineering workflows.
• A detail-oriented, systems-thinking mindset; you notice the one VCF out of ten thousand that doesn't look right.