Machine Learning Research Engineer

Machine Learning Research Engineer

Geometry, Gaussian Processes, Generative and Foundation Models

Location: Bay Area, California
Work style: In person or hybrid strongly preferred

About Expert Intelligence

OpenAI is building general AI for language and the internet.  Expert Intelligence (EI) is building the counterpart for scientific and analytical data. We focus on low data, high stakes domains such as pharma, food safety, and advanced materials. You will join a small Silicon Valley team of deep experts in ML, statistics, and scientific computing, and work on core models that sit at the heart of EI, similar to how foundation models sit at the heart of OpenAI.

Role Summary

As a Machine Learning Research Engineer you will

• Design and implement models that combine
  • manifold and geometric representations
  • Gaussian process and kernel methods
  • deep generative and foundation style models (transformers, diffusion, VAEs where relevant)
• Build training and evaluation pipelines in Python, JAX, and PyTorch for complex, structured and often low data scientific datasets
• Read recent research in geometry aware learning, GP based models, and generative AI, and turn ideas into working, efficient code
• Run experiments end to end, from data preparation to metrics and visualisation
• Collaborate with senior researchers to shape the next generation of EI models

This role is ideal for recent PhDs, strong MSc students, and exceptional undergraduates who enjoy both deep math and real code.

What We Look For

You do not need every item, but you should feel comfortable with most of them.

Mathematics and statistics

• Strong foundation in linear algebra, multivariate calculus, probability and statistics
• Comfort with proofs and abstract reasoning
• Exposure to at least one of
  • real or functional analysis
  • differential geometry or manifold theory
  • spectral methods or graph based learning
• Math Olympiad or similar competition background is a strong plus

Machine learning and generative AI

• Good understanding of supervised and unsupervised learning, overfitting, generalisation, and evaluation
• Experience or strong interest in at least one of
  • manifold learning or nonlinear dimensionality reduction
  • Gaussian processes or kernel methods
  • deep generative models such as VAEs, diffusion models, normalising flows, or transformers and large language models
• Ability to read research papers and implement algorithms from them

Programming

• Fluency in Python
• Experience with at least one modern ML framework, ideally PyTorch or JAX
• Ability to implement algorithms yourself, not only call high level library functions. Clean coding style and basic use of version control

Undergraduate and junior candidates are welcome if they can show deep mathematical understanding and strong coding skills through projects, theses, or competitions.

Compensation and Equity

This role is structured closer to a research fellow position than a senior big tech engineer role, with meaningful ownership in a high growth startup.

• Base salary, Bay Area
  • Approximately USD 95k to 135k per year
  • Strong undergraduates and MSc candidates are typically in the lower part of the range
  • Recent PhDs or candidates with exceptional competition or research track records are typically toward the upper part of the range
• Bonus and equity
  • Annual performance related bonus
  • Stock options or restricted stock representing real ownership in a high growth, high upside startup
  • Target total compensation at start is typically USD 120k to 170k per year, with room to grow as impact and responsibility increase
     

Key Benefits

• Health coverage
  Comprehensive medical coverage, with dental and vision included or available.
• Learning, conferences, and growth
  Budget for books, online courses, workshops, and conferences in ML, statistics, geometry, and AI, plus regular internal paper discussions and mentoring.
• Student loan and education support where possible
  For candidates with significant education costs, we aim to explore options such as partial student loan support, tuition for highly relevant courses, or extra support for conference travel tied to your work.
• Relocation and commuting support
  For candidates moving to the Bay Area we can discuss relocation support. For local staff we may support commuting costs consistent with other Bay Area tech employers.

Please send your CV or resume and a short note describing your background in mathematics and machine learning, one project where you implemented a nontrivial algorithm yourself, ideally involving geometry, Gaussian processes, or generative models.