Jua is working towards a universal physics simulator through a deep exploration of nature and the universe, their interaction with humans, business and the built environment. With a universal physics simulator, we collapse that barrier and help to use natural resources efficiently and sustainably whilst increasing human standard of living and spurring economic growth.
By joining our talented multidisciplinary team, you'll help shape our culture of ambition, transparent communication, and rapid iteration. We offer exciting challenges, creative freedom, fair pay, and generous stock options.
What we are looking for
As a Senior AI Researcher, you will develop and train state-of-the-art foundational earth systems models. As part of our fast-moving model research team, you will be working on developing and improving our deep learning-based algorithms and have great creative freedom and influence. Based on the latest publications, you will test and implement state-of-the-art research and, together with the team, build robust and sustainable machine learning pipelines for live operation.
Responsibilities and tasks
- Researching, implementing, and productionizing machine learning models for renewable power forecasting using satellite imagery, and other multi-modal data.
- Designing and training models on our GPU cluster with an emphasis on terabyte-scale environmental datasets.
- Creating novel algorithms to address challenges in climate modeling, from ideation and prototyping to live deployment.
Need to have
- Expertise in Python with the ability to quickly learn new programming languages as needed.
- Proficient in machine learning frameworks (e.g., PyTorch, TensorFlow, Keras, FLAX, TFX) and able to stay up to date with emerging tools and frameworks.
- Strong grasp of machine learning optimization techniques, including gradient clipping, mixed precision training, and layer parallelism
- Extensive knowledge of transformer architectures (e.g., GPT, Bart, Swin Transformer)
- Familiarity with advanced techniques such as diffusion models or graph neural networks and their relevance
- Experience with distributed training and inference architectures, including frameworks like Huggingface Accelerate and Microsoft DeepSpeed.
- Strong understanding of high-performance computing (HPC) and its role in accelerating machine learning workflows.
- Proven ability to work with terabyte-scale datasets, particularly those related to climate and geospatial data.
- General distributed systems knowledge and familiarity with public cloud offerings for ML workflows (e.g., GCP, AWS).
- Experience with tools like Git, Docker, Xarray, and Zarr.
- Demonstrated ability to independently solve complex problems with creativity and precision.
- Collaborative mindset with experience working in interdisciplinary teams.
- Awareness of the real-world, customer-facing impact of machine learning solutions, particularly in environmental applications.
- Background in GIS-based data, such as satellite imagery, radar data, or radiosondes.
Nice-to-have
- Experience with experiment tracking tools (e.g., WandB, MLflow).
- Hands-on experience running machine learning workloads in production environments.
- Contributions to open-source projects or the broader climate-focused machine learning community.
At Jua, we foster a performance culture and value people who embody our beliefs of service and adventure. We prioritize agility, operating at the highest clock speed to adapt quickly to change. We innovate on behalf of our users and leverage data supremacy to maintain our competitive edge. Through clear communication and fact-based decision-making, we ensure alignment in our pursuit of excellence. With these principles, we aim to create a customer-focused, value-centric organization that sets new standards in the industry. We value the unique perspectives that each individual brings to the table and believe that embracing diverse backgrounds and experiences enriches our collective journey towards growth and success.
Similar Jobs
.png){kind=logo}
What We Do
Jua.ai - Accuracy beyond expectation. Use beyond belief. The world's first end to end deep learning model for weather forecasting, delivered on an intuitive platform for anyone to customise it. Startups, governments and large companies alike can now develop global, high frequency and high accuracy weather models within days. Jua’s mission is focused on achieving artificial general intelligence (AGI) through a deep exploration of physics, the universe, and their interaction with the human civilisation. We are initially building the world's largest AI weather forecasting model so far and - to our knowledge - the first truly end to end one. This will help energy companies deal with weather volatility by way of significantly better and faster prediction as well as much more flexible extrapolation of insights.
