Remote AI Biologist: Spatial Genomics Benchmarks

LatchBio; San Francisco (CA); Hybrid; USD 120,000 - 180,000; Full time

Benefits offered by this job

2x free daily meals

Waterfront office near SF

Monthly parking

Job summary

LatchBio is seeking an Bioinformatics Engineer to join the Spatial team, testing AI agents on spatial transcriptomics projects and ensuring ripple effects are understood, not just mechanics. You’ll work at the frontier of AI in biology from a flexible, output-driven environment.

You’ll evaluate tissue alignment, segmentation, and deconvolution decisions, drawing from published papers and real data, with cross-domain learning and interoperability across teams.

Qualifications

• Must understand how image preprocessing affects segmentation and downstream analysis.
• Know trade-offs between spatial platforms and deconvolution vs segmentation.
• Be able to learn spatial tools and image-processing pipelines as needed.
• Design multi-checkpoint benchmarks and defend assumptions in variable tissues.

Responsibilities

• Design tasks that test AI agents on real spatial transcriptomics projects.
• Grade rationale behind segmentation thresholds and deconvolution choices.
• Build benchmarks handling registration failures and variable tissue morphology.

Skills

Systems thinking

Domain judgment

Tool flexibility

Task design

Tools

Baysor

STdeconvolve

giotto

Squidpy

Job description

LatchBio is seeking an Bioinformatics Engineer to join the Spatial team, testing AI agents on spatial transcriptomics projects and ensuring ripple effects are understood, not just mechanics. You’ll work at the frontier of AI in biology from a flexible, output-driven environment.

You’ll evaluate tissue alignment, segmentation, and deconvolution decisions, drawing from published papers and real data, with cross-domain learning and interoperability across teams.