ML Scientist

Harnham; San Francisco (CA); On-site; USD 150,000 - 230,000; Full time

Job summary

Harnham is seeking a Machine Learning Research Scientist to advance next-generation foundation models for complex multimodal datasets. This individual contributor role suits an independent researcher who designs rigorous experiments, builds models, and translates findings into actionable insights.

You will lead projects from concept to conclusion, collaborate with domain experts, and contribute to publications and conference submissions.

Qualifications

• Strong record of original machine learning research (conference or industrial).
• Doctoral-level research in ML or a quantitative discipline.

Responsibilities

• Design, train, and evaluate large-scale foundation models across multimodal datasets.
• Develop methods for combining information from multiple data sources, modalities, and abstraction levels.
• Create evaluation frameworks, benchmark tasks, and metrics to assess model effectiveness.
• Prototype research concepts quickly and assess potential impact.
• Collaborate with domain experts and ML practitioners to solve complex problems.
• Apply emerging AI approaches to research workflows and communicate findings to diverse audiences.
• Contribute to publications, conference submissions, and broader research engagement.
• Lead projects from concept formulation through experimentation and final conclusions.

Skills

Identifying meaningful problems

Designing rigorous experiments

Building models

Translating findings into insights

Education

PhD in Computer Science or related quantitative field

Job description

Machine Learning Scientist

Overview

A research-driven organisation is seeking a Machine Learning Research Scientist to help develop next-generation foundation models for complex scientific and multimodal datasets. This individual contributor position is ideal for an independent researcher who excels at identifying meaningful problems, designing rigorous experiments, building models, and translating findings into actionable insights. The role combines advanced machine learning research with practical model development. Success in this position will be measured primarily by scientific contribution, research quality, and the ability to generate new insights through experimentation.

Key Responsibilities

• Design, train, and evaluate large-scale foundation models across diverse multimodal datasets.
• Develop innovative methods for combining information from multiple data sources, modalities, and levels of abstraction.
• Create evaluation frameworks, benchmark tasks, and performance metrics to assess model effectiveness.
• Rapidly prototype research concepts and determine which ideas have the highest potential impact.
• Partner with domain experts, quantitative researchers, and machine learning practitioners to solve complex problems.
• Assess and apply emerging AI approaches, including large language models and autonomous agent-based systems, to research workflows.
• Communicate technical findings to both specialist and non-specialist audiences.
• Contribute to scientific publications, conference submissions, presentations, and broader research engagement activities.
• Lead projects from concept formulation through experimentation, analysis, and final conclusions.

Required Qualifications

Research Excellence

Candidates should demonstrate a strong record of original machine learning research through one or more of the following:

• First-author publications at leading machine learning, artificial intelligence, computer vision, or natural language processing conferences.
• Significant research contributions within highly regarded industrial research environments.
• Doctoral-level research in machine learning, computer science, statistics, artificial intelligence, or another rigorous quantitative discipline.

Key evaluation criteria

Key evaluation criteria include:

• Ability to identify important research questions.
• Technical rigor and depth of research execution.
• Evidence of innovative and independent thinking.

Execution Skills

The ideal candidate can:

• Independently move from concept to experiment.
• Implement and train machine learning models.
• Design robust evaluation methodologies.
• Perform comprehensive analysis of experimental outcomes.
• Iterate quickly based on evidence and results.
• Drive complete research efforts rather than contributing only a small portion of a larger initiative.

Academic Background

Strongly preferred:

• PhD in Computer Science, Machine Learning, Artificial Intelligence, Statistics, Applied Mathematics, Computational Neuroscience, Physics, or another quantitative field.

Candidates without a PhD may be considered if they possess an exceptional research record.

Technical Expertise

High-Priority Areas

• Foundation Models
• Self-Supervised Learning
• Representation Learning
• Computer Vision
• Multimodal Learning
• Large Language Models
• Generative AI
• Diffusion Models
• Flow-Based Generative Modelling
• Autoregressive Models
• Scientific Machine Learning

Additional Relevant Experience

• Robotics
• Autonomous Systems
• Computational Biology
• Molecular Modelling
• Protein Structure Prediction
• Molecular Interaction Prediction
• Structural Biology
• Biochemistry
• Quantitative Physics
• Astrophysics
• Other mathematically intensive scientific disciplines

Preferred Qualifications

• Experience applying machine learning to scientific or research-focused domains.
• Familiarity with computational life sciences, chemistry, drug discovery, or related fields.
• Demonstrated interest in using AI to accelerate scientific discovery.
• Strong written and verbal communication skills with the ability to explain complex concepts to diverse audiences.

Domain Knowledge

Prior domain expertise in life sciences is not required. The organisation prioritises machine learning excellence and research capability. Strong candidates may come from areas such as computer vision, language modelling, robotics, reinforcement learning, autonomous systems, or scientific AI.

Candidates should be able to articulate their interest in applying machine learning techniques to scientific challenges and emerging research opportunities.