Atmospheric Scientist

Gridmatic Inc. is seeking an Atmospheric Scientist to enhance their models and strategies for clean energy transition. This remote role involves integrating atmospheric data into prediction systems, collaborating with ML and engineering teams, and developing weather models. Responsibilities include evaluating weather products, running AI forecasts, and improving time series models. The position emphasizes teamwork, continuous learning, and the application of scientific knowledge in a fast-paced startup environment.
The Company
Gridmatic Inc. is a high-growth startup with offices in the Bay Area and Houston that is accelerating the clean energy transition by applying our expertise in data, machine learning, and energy to power markets. We are the rare startup that has multiple years of profitability without raising venture capital. Gridmatic is a great place to work with a culture that values teamwork, continuous learning, diversity, and inclusion. We move quickly and fix things. We are environmentally and data-driven, with a growth-oriented, academic mindset. We value integrity as much as excellence.

The Role
We are looking for an Atmospheric Scientist to apply their deep expertise to directly influence our models and strategies, contributing to the clean energy transition. This is a hands-on, individual contributor role where you will leverage your scientific knowledge and technical skills in a fast-moving, impactful startup environment. You will explore, evaluate, and integrate complex atmospheric data and models into our prediction systems, which drive real-time energy trading and optimization decisions. You will collaborate closely with our ML, engineering, and data science teams, teaching others about weather phenomena while learning about grid power dynamics and time-series modeling techniques.

What you might work on:
Research and develop & Internal Weather Models:
- Engaging in more open-ended research and development to build or refine our own weather modeling capabilities.
- Fine-tuning existing state-of-the-art AI models (e.g., based on GenCast, AIFS, and NeuralGCM).
- Post-processing existing SOTA AI forecasts to debias and recalibrate for our downstream power predictions.
- Incorporating and evaluating model changes, pushing the boundaries of how we forecast weather variables relevant to the energy sector.
- Educate and inform the broader team about atmospheric phenomena and weather forecasting concepts.
- Potential to publish research and findings derived from your work, contributing to the scientific understanding at the intersection of atmospheric science and energy, where appropriate and aligned with business goals.

Evaluating & Integrating External Weather Products:
- Surveying and evaluating the suitability of various Numerical Weather Prediction (NWP) and commercially available AI weather forecast products for our power production and price models.
- Rigorously evaluating and monitoring the performance of integrated weather products, analyzing their impact across different regions, timeframes, and weather regimes.
- Working with external data providers (like NOAA) and internal engineers to define data requirements.
- Work with engineers to build, monitor, and maintain data ingestion pipelines.

Evaluating & Running AI Weather Models In-house:
- Develop evolving metrics for AI weather models for our unique specifications.
- Set up, run, and monitor SOTA AI weather forecasts on our GPU cluster.

Generic Time Series Modeling:
- You might also apply your modeling skills to improve generic time series models for power production or energy price forecasting, using ML libraries like PyTorch or JAX.

Across all workstreams, you will be expected to:
- Write and maintain significant Python code within a Git-based software development workflow.
- Continuously learn about grid power modeling and the intricacies of energy markets.

You might be a good fit if you:
• Have earned a PhD in Atmospheric Science, Meteorology, or a closely related quantitative field.
• Have significant experience working with atmospheric models (NWP and/or AI models) and large meteorological datasets.
• Possess strong proficiency in Python programming.
• Some experience in Machine Learning with a desire to become an expert.
• Are comfortable working in a Linux environment and using Git for version control.
• Have experience running computational jobs on clusters or cloud computing environments.
• Have a demonstrated ability to analyze and interpret complex scientific data and model outputs.
• Are an effective communicator, able to explain complex weather concepts to non-experts.
• Are naturally curious and eager to learn about new domains, particularly energy systems and time-series modeling.
• Thrive in a fast-paced, dynamic startup environment where priorities can evolve

FAQ

What’s your policy on remote work?
We value the ability to work and collaborate in-person in our early stage as a startup, so Gridmatic has a hybrid policy of "50% in-office”. Most of the company works in our Cupertino office 2 or 3 days a week.

Join our team and make a difference! Click below or email us at careers@gridmatic.com.