AI Scientist

Nabla Bio, Inc; Cambridge (MA); On-site; USD 150,000 - 230,000; Full time

Benefits offered by this job

Equity

Benefits package

Job summary

Nabla Bio, Inc. is seeking an exceptional AI Scientist to lead development of core biomolecular modeling technologies in Cambridge, MA. You’ll build and improve foundation models powering therapeutic design, with architecture design, training, and experimental validation.

You will work closely with wet-lab scientists and software engineers to productionize models for internal and pharma partner workflows, leveraging large-scale experimental feedback to drive state-of-the-art performance.

Qualifications

• 5+ years of experience developing deep learning models.
• Experience with generative modeling or protein/biomolecular ML is a plus.
• Familiarity with large-scale sequence modeling or HPC/cloud training.
• Strong engineering fluency and ability to lead ML research efforts.

Responsibilities

• Design, implement, and evaluate new training data, model architectures, training schemes, and loss functions for biomolecular generation and prediction.
• Drive major improvements in generative and predictive performance based on experimental feedback.
• Collaborate with AI engineers to productionize models for internal and pharma partner workflows.
• Stay on top of state-of-the-art ML, protein modeling, and sequence design—and push it forward.

Skills

Deep learning

Python

PyTorch

Distributed training

ML research leadership

Tools

Python

Job description

The Role

We’re hiring an exceptional AI Scientist to lead development of our core biomolecular modeling technologies. You’ll be responsible for building and improving the foundation models that power Nabla’s therapeutic design capabilities — from architecture design and training to experimental validation.

This is a rare opportunity to do AI research with real-world, large-scale experimental feedback: your models will be tested not just with loss curves and benchmarks, but in wet-lab assays measuring therapeutic function, safety, and precision. Our platform enables you to test dozens of modeling hypotheses in parallel, with experimental results across a million drug designs returned in just a few weeks. See our papers for examples of our work [1][2], and their coverage in Science Magazine and Endpoints News.

This is an in-person role in Cambridge, MA. You will:

• Design, implement, and evaluate new training data, model architectures, training schemes, and loss functions for biomolecular generation and prediction
• Drive major improvements in generative and predictive performance based on experimental feedback
• Collaborate with AI engineers to productionize models for use in internal and pharma partner design workflows
• Stay on top of the state of the art in ML, protein modeling, and sequence design—and push it forward

Qualifications

• 5+ years of experience developing deep learning models; prior experience in generative modeling, protein/biomolecular ML, or large-scale sequence modeling is a plus
• Strong engineering fluency in Python and PyTorch
• Experience with distributed training and scaling large models in HPC/cloud environments
• Track record of creativity, rigor, and technical leadership in ML research
• Comfort working closely with experimentalists to connect model behavior to real biological outcomes

What We Offer

• The ability to test and validate ML hypotheses using one of the most powerful experimental platforms in biotech
• A chance to shape foundational modeling capabilities for programmable drug design
• Close collaboration with experts in wet-lab biology, bioinformatics, and software engineering
• A focused, technically ambitious team solving hard problems end-to-end
• Highly competitive salary, equity, and benefits package