About Prima Mente

• Implement high-performance ML algorithms optimised for massive-scale, ensuring reliability, efficiency, and scalability.
• Design, develop, and maintain robust experimentation pipelines enabling rapid iteration, precise evaluations, and reproducible research outcomes
• Refactor and scale prototype research code into clean, maintainable, and performant repositories suitable for production-grade deployments.
• Create high-speed data processing workflows capable of efficiently handling large-scale datasets to accelerate experimentation and model development.
• Experimental design, prioritising high impact experiments with the highest signal:noise ratio.

• In 1 month you will be responsible for running initial experiments with state-of-the-art machine learning models, reviewing and implementing cutting-edge research papers, and optimizing existing code for efficiency and accuracy.
• In 3 months you will directly own and have created a prototype model architecture, demonstrated significant algorithmic improvements, and contributed to scaling methods for large-scale data ingestion and training.
• In 6 months, you’ll have developed a high-performance version of a foundation model, implemented key algorithmic optimizations that boost scalability and throughput, and published internal benchmarks demonstrating significant research impact.

• Meaningful Impact: Contribute directly to research infrastructure that powers discoveries potentially impacting millions of lives.
• Innovation & Autonomy: Work at the forefront of AI and multi-omics, with the freedom to propose and implement state-of-the-art infrastructure solutions.
• Exceptional Team: Collaborate with talented colleagues from diverse backgrounds across ML, bioinformatics, and engineering.
• Growth Opportunities: Continuous learning and growth opportunities in a rapidly advancing technical field.

• Exceptional performance at exceptional pace
• The solutions we build demand uncompromising quality and rigour.
• The problems we are solving are grave and present.
• Inquisitive discovery
• We embrace curiosity and creativity.
• Every question is a path to a transformational breakthrough.
• Radical candour
• We practice unwavering honesty and transparency in all our challenges and interactions.
• Purposeful individuality
• Every individual in our team is celebrated for their identity, uniqueness, and experiences.
• We are invested in each one’s bespoke personal development.
• Nurturing individuality will supercharge our collective purpose and spirit.
• Patient impact at scale
• We have a steadfast commitment to improve the health and well-being of patients globally.
• Every experiment run, every dataset analysed, and every innovation developed, is a step towards achieving a scalable impact.
  

• Ambitious and Impact-Driven: You're inspired by working at the forefront of AI and biology, motivated by challenges that can significantly advance human health.
• Technical Excellence: You thrive in highly technical, complex environments and have a track record of turning cutting-edge research into robust production systems.
• Collaborative & Communicative: You excel at collaborating across disciplines, clearly articulating complex ideas, and driving alignment among research and engineering teams.

• Deep understanding of state-of-the-art machine learning methodologies and proven experience in translating them into practical solutions.
• Solid foundation in distributed computing principles, parallel processing, and algorithmic efficiency.
• Experience optimizing ML algorithms for performance, memory efficiency, and compute resource utilization.
• Skilled in designing and implementing scalable data pipelines capable of rapid ingestion, transformation, and processing.
• Deep expertise in modern ML frameworks and tools (e.g., PyTorch, JAX, TensorFlow), and familiarity with state-of-the-art training and inference workflows.
• Skilled in clearly articulating complex ideas, effectively communicating why particular approaches succeed or fail, and providing insightful critical analyses.
• Experience of building highly collaborative research teams.
• Track record of working on hard problems for long periods of time.
• High agency with the ability to jump on any task as needed.
• Demonstrated experience training, optimizing, and deploying large-scale models (7B+ parameters).
• Low level algorithm optimisation
• Quantization (8bit or lower)
• JIT compilation
• XLA/Mosaic/Triton/CUDA
• Hardware optimisation (GPU/TPU/HPU)
• Finetuning Optimization (QLora, QDora)
• Large scale data above 2T tokens