Machine Learning Researcher

Join to apply for theMachine Learning Researcherrole atFoster + Partners.

Foster + Partnersis a global studio for architecture, engineering, urban and landscape design, rooted in sustainability.

This role will include strategic R&D around developing, implementing and evaluating new models, agents and techniques, as well as deploying standard models and methods towards prediction and automation - including building the relevant software to support wider deployment of these derivative applications. The role will include researching the latest models in Machine Learning and implementing proofs of concept for problems particular to the AEC industry, using standard frameworks such as PyTorch and JAX.

Once a model is fully tested, the role will involve deployment to test and production on premises or the cloud. The role also involves identifying and leveraging opportunities in ML research for both architectural design innovation and operational business applications, training and grounding image and language-based ML models on proprietary data and investigating workflow automation through AI agents. You will work alongside data scientists, front-end and back-end developers as well as domain specialists like material scientists and business analysts.

Responsibilities

• Master’s degree in Computer Science, Machine Learning, Mathematics or equivalent experience.
• Deep knowledge and understanding of Deep Learning architectures.
• Ability to implement latest research ML papers.
• Strong programming skills, with a preference for Python.
• Excellent experience developing and training deep learning models using PyTorch and/or JAX.
• Excellent experience with Sci-Kit Learn, Pandas and Numpy.
• Experience with CUDA or Triton.
• Experience with Computer Vision, Natural Language Processing and Graph Neural Networks.
• Experience with a wide range of generative modelling techniques, including diffusion based models, flow matching, and Latent Consistency Models for image synthesis, as well as large-scale transformer architectures for language modelling, with a focus on underlying algorithms and emerging industry trends.
• Knowledge of DataOps and MLOps concepts and tools (e.g. MLFlow).
• Experience with Huggingface libraries like Transformers, Diffusers and Accelerate.
• Effective communication skills, with the ability to deliver compelling presentations on how Machine Learning models can help the business.

In return we offer a competitive basic salary and generous benefits package which includes 25 days holiday (exc bank holidays), Pension, DIS and discretionary annual bonus.

Employment Details

• Seniority level:Not Applicable
• Employment type:Full-time
• Job function:Research
• Industries:Architecture and Planning

#J-18808-Ljbffr