Senior ML Scientist: Multimodal & Relational Foundations

Altos Labs; San Diego (CA); Hybrid; USD 179,000 - 307,000; Full time

Job summary

Altos Labs seeks a Staff Machine Learning Scientist to advance unified, multi-modal foundation models for aging and disease research. You will design, implement, and scale state-of-the-art models across Institute of Computation teams, focusing on synthesis of unstructured signals with relational data.

The role emphasizes collaboration, scientific rigor, and contributions to publications, while mentoring junior staff and driving production-ready research into scalable systems.

Qualifications

• PhD in Computer Science, ML, or similar field; 5+ years of relevant work.
• Experience in developing novel generative AI models, multimodal, GraphRAG, or relational DL.
• Strong understanding of ML principles and architectures (Transformers, GNNs, diffusion).
• Strong Python programming and deep learning libraries experience (PyTorch, JAX, HF).
• Experience with multi-GPU distributed training (DDP, FSDP, DeepSpeed, Megatron, Ray).
• Proven publishing track record in top AI/ML conferences.

Responsibilities

• Pre-train and fine-tune large-scale ML systems using multimodal data.
• Architect novel hybrid models combining LLMs and GNNs for reasoning over biological graphs.
• Develop Relational Foundation Models for zero-shot tasks on diverse datasets.
• Lead data loading strategies and distributed training recipes across GPUs.
• Apply theory and math to evaluate model performance and scalability.
• Mentor junior staff and contribute to peer-reviewed publications.

Skills

PhD in CS/ML

Generative AI

ML architectures

Python & DL libs

Distributed training

Publications

Tools

PyTorch

JAX

Hugging Face

Job description

Altos Labs seeks a Staff Machine Learning Scientist to advance unified, multi-modal foundation models for aging and disease research. You will design, implement, and scale state-of-the-art models across Institute of Computation teams, focusing on synthesis of unstructured signals with relational data.

The role emphasizes collaboration, scientific rigor, and contributions to publications, while mentoring junior staff and driving production-ready research into scalable systems.