Machine Learning Research Engineer, Scientific AI

Just Food Company•Emeryville, CA

8d•$160,000 - $200,000•Onsite

About The Position

Just Food Company is seeking a Machine Learning Research Engineer to apply machine learning to scientific research and practical company-wide applications. The primary focus is developing systems that help identify, evaluate, and validate proteins with valuable food functionalities. The role aims to create a faster cycle between prediction and experimentation: selecting promising proteins, predicting their performance in food, choosing informative experiments, and improving future predictions and decisions based on experimental results. In R&D, the engineer will develop models connecting protein sequence, structure, biochemical properties, formulations, and experimental data to food functionality. They will work closely with protein scientists, food scientists, and product developers to guide protein selection and experimentation. Potential methods include protein representation learning, graph and geometric deep learning, multimodal and generative models, active learning, and Bayesian optimization, selected according to the problem and evaluated against real experimental outcomes. The role also applies AI across Just Food Company. The engineer will work with teams such as product development, sales, and operations to identify valuable opportunities, rapidly prototype and evaluate solutions, and turn promising prototypes into reliable tools. This hands-on position offers significant ownership and autonomy and involves collaboration with scientists, functional teams, and company leadership. The position is full-time and regular, based onsite five days per week at the company headquarters in Emeryville, California. The expected base salary range is $160,000–$200,000 per year. Candidates must be authorized to work in the United States without current or future employer sponsorship.

Requirements

Strong foundations in machine learning and demonstrated experience developing modern machine learning models and systems.
Excellent Python skills and hands-on experience with PyTorch, JAX, TensorFlow, or a comparable machine learning framework.
Experience taking machine learning problems from data and experimentation through evaluation and implementation.
Ability to reason across data, model architecture, training, evaluation, inference, and deployment.
Ability to select methods based on the structure of a problem rather than defaulting to a particular model or technology.
Comfort working with sparse, noisy, heterogeneous, or partially observed data.
Strong software engineering fundamentals and commitment to testing, reproducibility, maintainability, and performance.
Ability to balance open-ended research with practical implementation and iteration.
Intellectual curiosity and ability to quickly develop fluency in unfamiliar scientific and business domains.
Strong communication skills and enthusiasm for working directly with scientists, product developers, operators, commercial teams, and company leadership.
Comfort operating with significant ownership and autonomy in a small, fast-moving organization.
Candidates must be authorized to work in the United States without current or future employer sponsorship.

Nice To Haves

Experience with protein representation learning, protein language models, graph or geometric deep learning, diffusion or other generative models, multimodal learning, active learning, or Bayesian optimization.
Experience with biological sequences, protein structures, molecular graphs, biochemical measurements, or other structured scientific data.
Experience in computational biology, bioinformatics, protein engineering, structural biology, chemistry, food science, ingredient discovery, or another scientific machine learning domain.
Experience designing computational approaches alongside laboratory scientists and incorporating experimental results into subsequent modeling.
Experience with language-model applications, retrieval systems, tool-using agents, or internal AI applications.
Experience with data pipelines, GPU workloads, cloud infrastructure, distributed training, or scientific computing.
Research, open-source projects, publications, patents, datasets, or production machine learning systems demonstrating exceptional technical ability.
A BS, MS, or PhD in Computer Science, Machine Learning, Computational Biology, Bioinformatics, Applied Mathematics, Statistics, or a related quantitative field is relevant; demonstrated ability is valued over a particular degree or academic pedigree.

Responsibilities

Develop models that predict food functionality from protein sequence, structure, biochemical and biomaterial properties, formulations, and experimental data.
Build representations and models capturing relationships among proteins, ingredients, formulations, processing conditions, and functional outcomes.
Select and apply appropriate approaches, potentially including protein representation learning, graph and geometric deep learning, generative and diffusion models, multimodal learning, active learning, and Bayesian optimization.
Partner with experimental scientists to design studies, evaluate predictions, and incorporate experimental results into subsequent modeling.
Build data pipelines, training infrastructure, evaluation frameworks, and research tools for iterative model development.
Develop prospective evaluations to determine whether model predictions improve protein selection and experimental decision-making.
Translate successful research into practical tools for scientists and product developers.
Partner with product development, sales, operations, and other teams to identify high-value AI opportunities.
Build practical AI systems that improve knowledge access, analysis, planning, reporting, and execution.
Develop applications and agents that use company data, knowledge, and systems to perform useful work.
Prototype and evaluate potential solutions, then turn the strongest ideas into reliable tools.
Apply appropriate permissions, testing, human oversight, and measurement to deployed systems.
Help establish shared infrastructure and development practices for reliable company-wide AI development and use.
Measure outcomes such as adoption, time saved, decision quality, scientific impact, and work completed.