Consultant in Meteorology and Machine Learning

AI Weather Forecasting Expert – Technical Due Diligence on Acquisition Target (based in China) Project Description We're evaluating the acquisition of an AI meteorology company. We're looking for an independent expert in machine learning for weather forecasting to conduct confidential technical due diligence on the target's AI forecast model and MLOps system, and deliver a clear, evidence-based key findings report. You'll work directly with designated representatives on our side. All contact with the target will be coordinated through us. Scope of review 1. Data acquisition – sources, licensing, quality and pre-processing of training and inference data 2. Model architecture – design, training methodology, comparison with the state of the art weather forecasting models (e.g. Google's Weathernext3) 3. Model output – resolution, lead times, calibration, uncertainty quantification, performance in extreme events 4. MLOps system – pipelines, versioning, deployment, monitoring, retraining, infrastructure, technical debt 5. Benchmarking – independent evaluation of the target's forecast skill against an agreed list of reference systems, verification datasets and metrics Deliverables 1. DD Plan: a request list of documents, data, code, access and interviews needed from the target; a Benchmark List (reference systems, datasets, metrics, variables, lead times, regions); and a work plan with estimated hours and dates 2. Key Findings Report (draft then final): covers all five areas including benchmarking methodology and results; ranks findings as Critical / Major / Minor with evidence and recommended actions; notes any information the target did not provide and the resulting limitations 3. All supporting scripts, benchmark results and notes. 4. Immediate escalation of any material risk or "red flag" as soon as identified Requirements 1. PhD (or equivalent research experience) in ML, atmospheric science, meteorology or a related field Hands-on experience with AI weather models (e.g. GraphCast, Pangu-Weather, FourCastNet, GenCast, FuXi, AIFS) and familiarity with NWP baselines (ECMWF IFS, GFS) 2. Strong forecast verification skills – ERA5, WeatherBench 2, metrics such as RMSE, ACC, CRPS, extreme-event skill 3. Python with PyTorch or JAX; working knowledge of MLOps practices (experiment tracking, versioning, CI/CD, monitoring) 4. Excellent written English and ability to explain technical findings to a business audience 5. Preferred: prior technical due diligence, code audit or expert review experience; publications in AI weather/climate; Asia time-zone availability To apply, please include: 1. A short summary of your AI weather forecasting and verification experience 2. One or two examples of benchmarking, audit or review work you've done 3. Your hourly rate and weekly availability over the next 2–3 months