Overview

Jua is on a mission to achieve artificial general intelligence (AGI) by delving deeply into physics, the universe, and their relationship with human civilization. We are operating a foundational earth systems model. Our first model is capable of predicting atmospheric physics globally, with state-of-the-art performance in comparison to existing weather models.

As we scale the model, more applications like wildfire, flood or vegetation prediction become possible. Our flagship product is an API designed for weather-dependent power and energy traders, offering superior accuracy, lower latency, and a higher refresh rate, especially during critical events like hurricanes and cyclones.

With substantial venture capital backing at a sensible valuation, our first objective will be to provide businesses and countries with several orders of magnitude better and faster weather forecasts. Our multidisciplinary team combines expertise in machine learning, physics, aerospace, data processing, and UX design to create a crucial component for a sustainable world amidst changing climate patterns and increasing extreme weather events.

Joining our team means collaborating with individuals who have built successful companies and worked on technologies used by millions. You'll contribute to our culture of ambition, transparent communication, rapid iteration, and humility. We offer exciting challenges, creative freedom, a talented and enthusiastic team, fair compensation, and generous shares in the company.

What we are looking for

As a Staff AI Engineer, you will help develop state-of-the-art foundational earth system models based on machine learning and open & proprietary data. As part of our fast-moving research team, you will be working on developing and improving our deep learning-based algorithms and have great creative freedom and influence. You will bring an engineering perspective to research driven development and play a leading role in the scaling of our models. You will understand and appreciate the tradeoffs and challenges associated with scaling models >1 billion parameters over hundreds of GPUs.

Responsibilities and tasks

• Developing and improving deep learning-based weather forecasts with state-of-the-art technologies
• Training these models on our GPU cluster
• Developing new algorithms from the idea through the prototype to live operation
• Testing, validating, visualizing, as well as documenting new approaches and their performance

Need-to-have

• You have at least 5 years of professional experience in Machine Learning (ignore this point if you can prove in another way to have exceptional skills in that regard)
• You are proficient in Python
• You are familiar with the common machine learning frameworks (PyTorch or FLAX)
• You are comfortable with linear algebra and have a clear understanding of the detailed processes going on in deep neural networks
• You are familiar with SOTA transformer architectures like GPT, Bart or Swin Transformer
• You have implemented at least one SOTA paper successfully from scratch by reading & understanding the paper
• You are familiar with common techniques like gradient clipping, layer parallelism or mixed precision to train very deep networks
• You are familiar with distributed training using both PyTorch build-ins as well as libraries like Huggingface Accelerate or Microsoft Deepspeed
• You understand how to approach scaling large models (e.g. >1 billion parameters) on distributed infrastructure including both multi-GPU and distributed data stores
• You have an intrinsic interest in improving our life on Earth with technology
• You have experience in handling and training with large amounts of data (>1TB)
• You like to move fast, break things, take risks, think out of the box, iterate & learn fast

Nice-to-have

• You are familiar with experiment tracking tools (like WandB or ML flow)
• You have experience with GIS-based data such as satellite, radar or radiosondes
• You have experience running Machine learning workloads in production environments

Even if you don't have all the qualifications but are enthusiastic about the job, we would be happy to get to know you! We believe that anyone interested in a topic can learn something new. So don't hesitate to apply.

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.