Machine Learning Researcher - Rainmaker Technology

About the Role

Rainmaker is hiring its first dedicated Machine Learning Researcher. You will establish how Rainmaker uses machine learning across the company: identifying the most valuable problems, determining which are ready for ML, building working models, and partnering with engineers and domain experts to turn successful research into operational systems.

You will not inherit a single predetermined model roadmap. The opportunity set includes forecasting supercooled liquid water and cloud-seeding opportunities, assimilating multimodal observations into estimates of atmospheric state, improving microwave-sounder retrievals, predicting hail, learning from intervention outcomes, and finding other high-leverage applications across research and operations.

Rainmaker's long-term advantage is not a generic weather model. It is the combination of proprietary in-cloud observations, radar and satellite data, UAS measurements, field campaigns, and repeated atmospheric interventions. You will build the learning systems that turn those data into better estimates, predictions, and decisions.

This is initially a hands-on individual-contributor role. You may eventually help recruit or technically lead an ML team if that fits your strengths and Rainmaker's needs, but management is not an initial expectation.

What You'll Do

• Assess potential ML projects across Rainmaker and prioritize them by operational value, data readiness, technical tractability, and time to useful results.
• Deliver an operationally useful model or prototype within your first three months rather than spending a quarter exclusively on infrastructure or roadmap development.
• Build models for forecasting, nowcasting, retrievals, multimodal atmospheric-state estimation, simulation, intervention analysis, and other scientific or operational applications.
• Develop methods for forecasting the occurrence, location, amount, and persistence of supercooled liquid water at scales relevant to cloud-seeding operations.
• Combine public NWP, radar, satellite, microwave-sounder, aircraft, UAS, sounding, surface, and in-situ observations.
• Build datasets, labels, baselines, evaluation metrics, and validation procedures for variables that public systems do not observe or optimize well.
• Establish honest experimental comparisons and characterize calibration, uncertainty, generalization, and failure modes.
• Work closely with meteorologists and atmospheric scientists to define targets, physical constraints, useful priors, and ground truth.
• Write research-quality software and build prototypes that software engineers can help productionize when an approach proves valuable.
• Use Rainmaker's compute budget deliberately, scaling experiments only when the problem, data, and baseline justify it.