Machine Learning Performance Engineer, London
London
Machine Learning Performance Engineer, London
We are here to advance human health by reimagining drug discovery with the power and pace of artificial intelligence.
The future is coming, enabled and enriched by the incredible power of machine learning. A future where diseases are curtailed or cured through faster and better drug discovery.
Our values serve this future. We believe they will help us bring it closer.
Join an interdisciplinary team driving groundbreaking innovation and contribute meaningfully to our ambitious goals, within an inspiring and collaborative culture.
The world we want tomorrow starts today, with our culture and with you.
About Isomorphic Labs
Founded in 2021 and led by Sir Demis Hassabis, Isomorphic Labs aims to usher in a new era of biomedical breakthroughs and find cures for devastating diseases.
Building on the success of Google DeepMind’s AlphaFold, we develop state-of-the-art technologies to accelerate and improve medicine design and delivery.
Our world-leading drug design engine uses foundational AI models across multiple therapeutic areas and drug modalities, continually innovating to advance rational drug design.
Your Impact
We seek engineers at Mid to Senior, Staff, or equivalent levels to shape the performance and scaling capabilities of IsoLabs.
What You Will Do
* Develop custom GPU kernels to maximize utilization and performance.
* Design, implement, and optimize distributed training and inference strategies.
* Implement low-level hardware optimizations.
* Design low-precision algorithms that balance performance and accuracy.
* Optimize performance for latency and throughput in real-world drug design programs.
* Collaborate with infrastructure teams to deploy solutions and ensure training uptime.
* Work with ML engineers and researchers to create efficient model architectures.
Skills and Qualifications
* Strong knowledge of HPC and ML systems.
* Understanding of GPU and AI accelerator architectures.
* Deep understanding of data structures and algorithms.
* Experience with deep learning frameworks, preferably JAX.
Nice to Have
* Knowledge of XLA, Triton, CUDA, Pallas, or similar DSLs / compiler stacks.
* Experience with distributed training and sharding strategies.
* Knowledge of collective communication libraries like NCCL.
* Proven ability to optimize ML accuracy with low-precision formats.
* Experience deploying and maintaining systems on GCP.
* Interest in chemistry and biology.
Culture and Values
Our shared values guide our work and strengthen our culture:
* Thoughtful: Curiosity, creativity, and care in rigorous science.
* Brave: Fearlessness, initiative, and integrity in facing challenges.
* Determined: Confidence, urgency, and agility to pursue our goals.
* Together: Collaboration and connection to create impact.
Building an Extraordinary Company
We value diverse skills and backgrounds, fostering an environment of collaboration, shared learning, and support. We are committed to equal employment opportunities and inclusive practices.
Our hybrid work model requires in-office presence 3 days a week (typically Tuesday, Wednesday, and one other day). If you have special needs regarding this approach, we are happy to discuss accommodations.
Privacy Notice: When applying, your data will be processed in accordance with our privacy policy. Accepted file types: pdf, doc, docx, txt, rtf.
#J-18808-Ljbffr