Pelmorex Corp is hiring a
Team Lead, AI Software Engineering

Summary

Pelmorex is seeking an experienced AI Software/Research Engineer with expertise in Deep Learning Weather Prediction (DLWP) to help transform their research organization. This remote position requires 7+ years of professional experience, at least 1 year of formal leadership experience, and a meteorological or AI background. The role involves ideating, project managing, and executing DLWP research tasks, mentoring team members, and helping the engineering organization transition to an 'AI first' approach.

Requirements

• 7+ years of professional experience
• At least 1 year of formal leadership experience, or clear demonstration of leadership interest and abilities, including strong interest in teaching and mentorship
• Meteorological and/or AI background: Coursework or published research in weather/climate forecasting, artificial intelligence, or both
• Hands-on AI experience: Experience with all parts of the AI lifecycle, including training and deploying data-hungry deep learning models in a commercial context
• Scientific Python: Xarray, Dask, Numpy, Scipy, Pandas, Scikit-learn
• ML Python: JAX, Haiku, PyTorch, TensorFlow, Caffe, Keras
• Low-level memory management frameworks: OpenMPI, CUDA, XLA
• Debugging skills: training and learning/loss dynamics, time/memory/utilization profiling, multi-process debugging, and log management
• Distributed training experience: muti-node, multi-accelerator environments
• Knowledge of modern DL architectures and how they apply to weather/climate prediction problems
• Generalist software engineering and cloud infrastructure skills: Coding, Cloud, Linux, Traditional NWP/HPC experience a plus
• Experience with weather and climate data: Grib, NetCDF, Arrow IPC, Parquet, Zarr and supporting libraries (Eccodes, Xarray, etc)
• Spatial and temporal interpolation techniques
• Knowledge of forecast and climate models, parameters, and verification techniques
• Exceptional leadership skills: Communication, presentation, project management, including technical writing and design skills
• Influential: leverages personal and team experience to guide decision making diplomatically and effectively
• Optimistic, open minded, inclusive

Responsibilities

• Ideate, project manage, and execute against DLWP research tasks
• Manage the ‘buy or build’ question as it relates to adopting and adapting open source DLWP models
• Train, verify and deploy DLWP models on enormous amounts of data using SOTA on-prem and cloud accelerators
• Champion technical designs, recruit feedback, code and configure, project manage, and document ML systems, data, and pre/post processing pipelines
• Mentor and lead meteorologists, climatologists, data scientists, and software engineers
• Help the engineering organization transition to ‘AI first’
• Bring a production software mindset to research and development tasks
• Conduct and solicit code reviews
• Create a culture of inclusivity, psychological safety, honesty, and vulnerability
• Coordinate roadmaps with key product and research stakeholders

Preferred Qualifications

• Think ‘at scale’: default to documentation, automation, source control, force multiplication, etc
• Have hands-on DLWP experience with one/some of: GraphCast, MetNet, ClimaX, FourCastNet, SEEDS, GenCast, Pangu Weather, FuXi, WeatherBench, CorrDiff, …

Benefits

• Remote Work Environment
• Personal Spending Account
• Up to $500 to spend on anything related to physical, mental or financial well-being (in addition to your Health Care Spending Account)
• Summer Hours: Start your weekend earlier during the summer months!
• 17 Paid Days Off (in addition to 13 Personal Days). This includes an extra day off during the month of your birthday - our gift to you!
• An IDEAS (Inclusion, Diversity, Equity, Awareness, Solidarity) team
• Weather is inclusive, we will be too. Our IDEAS team is committed to making this happen!
• Course Reimbursement Program
• Free online doctor visits with Maple Online Healthcare
• While we encourage 1:1 conversations, we recognize that not everyone is comfortable with speaking up. We have an anonymous reporting platform (Speakfully) to ensure everyone’s voice is heard