AI Scientist - Biomedical Multimodal Modeling

at Xaira Therapeutics

Location

South San Francisco, California, United States

Compensation

$170k–$240k USD

Type

full time

Posted

2 weeks ago

Save to queue

Market range · function + seniority

$73k—$123k—$175k

p25 · target · p75 · n=800

Posted $240k · well above market

Tailor your résumé to this role in 30 seconds.

Free account · ATS keyword check · per-job bullet rewrite by Claude.

Tailor my résumé Apply on company site

Job description

About Xaira Therapeutics

Xaira is an innovative biotech startup focused on leveraging AI to transform drug discovery and development. The company is leading the development of generative AI models to design protein and antibody therapeutics, enabling the creation of medicines against historically hard-to-drug molecular targets. It is also developing foundation models for biology and disease to enable better target elucidation and patient stratification. Collectively, these technologies aim to continually enable the identification of novel therapies and to improve success in drug development. Xaira is headquartered in the San Francisco Bay Area, Seattle, and London.

About the Role
We are building foundation models for biological systems, focusing on learning from complex, multimodal data. Our approach combines representation learning and generative modeling to capture structure, variation, and latent organization in high-dimensional settings. We work with both image-based data and high-dimensional molecular (omics) measurements, treating them as complementary views of the same underlying system. Our goal is to develop models that can both encode rich representations and generate or infer missing aspects from partial observations.

What You’ll Do

Develop multimodal foundation models spanning image and molecular data
Design approaches that unify representation learning and generative modeling
Train large-scale models on structured, high-dimensional datasets
Explore cross-modal alignment, prediction, and generation
Build scalable systems for data processing, training, and evaluation
Work in a research-driven environment with evolving problem definitions

What We’re Looking For

Strong background in deep learning, especially in modern architecture

Experience with representation learning (self-supervised, contrastive, masked modeling, etc.)
Experience with generative modeling (diffusion, autoregressive models, VAEs, or related approaches)
Proficiency in Python, ML frameworks, and CI/CD (PyTorch, JAX, etc.)

Relevant Experience

Work on multimodal learning or cross-domain representation learning
Experience with image data, molecular / omics-like data, or both
Familiarity with large-scale training

Bonus

Exposure to biological or scientific datasets
Experience with paired or partially aligned multimodal data
Comfortable working in open-ended, evolving problem spaces

Compensation

We offer a competitive compensation and benefits package, seeking to provide an open, flexible, and friendly work environment to empower employees and provide them with a platform to develop their long-term careers. A Summary of Benefits is available for all applicants. We offer a competitive package that includes base salary, bonus, and equity. The base pay range for this position is expected to be $170,000 - $240,000 annually; however, the base pay offered may vary depending on the market, job-related knowledge, skills and capabilities, and experience.

Xaira Therapeutics an equal-opportunity employer. We believe that our strength is in our differences. Our goal to build a diverse and inclusive team began on day one, and it will never end.

TO ALL RECRUITMENT AGENCIES: Xaira Therapeutics does not accept agency resumes. Please do not forward resumes to our jobs alias or employees. Xaira Therapeutics is not responsible for any fees related to unsolicited resumes.

More open roles at Xaira Therapeutics

Hiring velocity, headcount trend, and every open posting on one page.

View Xaira Therapeutics profile →

Similar roles

Open postings ranked by description similarity — useful if this role isn't quite right.