Staff AI Engineer - Multimodal Genomics & Biosecurity

Radical Numerics; San Francisco (CA); On-site; USD 180,000 - 260,000; Full time

Benefits offered by this job

Competitive compensation

Comprehensive benefits

Support for continual learning

Job summary

Radical Numerics is an AI research lab building general biological intelligence. We design and deploy multimodal genome language models to advance biosecurity and biodefense, while addressing real-world risks.

You’ll collaborate with computational biologists, AI scientists, and systems engineers to translate advances into deployable capabilities and security-focused solutions. We seek research scientists and engineers with a deep learning background applied to biological problems, strong Python

Qualifications

• Experience applying ML to biological problems.
• Strong Python and ML tooling experience.
• Familiarity with molecular biology and genomic data modalities.
• Ability to reason about real-world risks of AI in biology.

Responsibilities

• Build and fine-tune frontier AI models.
• Design experiments applying large models to biosecurity problems.
• Bridge to real-world biosecurity systems and deployable tools.
• Analyze model behavior and ablations.
• Explore mechanistic interpretability for sequence data.
• Prototype new approaches using foundation models.
• Operate with high ownership in a fast-moving environment.

Skills

Deep learning

Evaluation benchmarks

Python

ML tooling

Biological data ML

Risk assessment

High agency

Tools

PyTorch

JAX

Experiment management

Distributed training

Job description

Radical Numerics is an AI research lab building general biological intelligence. We design and deploy multimodal genome language models to advance biosecurity and biodefense, while addressing real-world risks.

You’ll collaborate with computational biologists, AI scientists, and systems engineers to translate advances into deployable capabilities and security-focused solutions. We seek research scientists and engineers with a deep learning background applied to biological problems, strong Python