LatchBio

Engineer

• You'll design, build, and evaluate LLM benchmarks for genomic analysis - spanning read alignment, assembly, variant calling, GWAS, causal inference, and multiomics integration.
• Your benchmarks and evaluations will set the bar for how well AI agents can serve as analysis partners for scientists working with human genetics data.
• You'll own the full evaluation pipeline: choosing datasets, defining failure modes, writing grading scripts, and shipping evaluations that measure real capability.
• Your work directly impacts how LLMs reason through genomic problems.

Bioinformatics Engineer – Variant