Founding Geospatial ML Engineer for Production Forecasts

Worldcastr; New York (NY); On-site; USD 310,000 - 335,000; Full time

Job summary

Worldcastr is seeking a founding Geospatial Machine Learning Engineer to turn our predictive spatial models into production-ready, validated forecasting capabilities. You will own end-to-end modeling across parcels, buildings, and regional indicators, building robust datasets and reproducible feature pipelines.

The role emphasizes probabilistic forecasting, geographic transfer, explainability, and rigorous evaluation across horizons, with a strong emphasis on production delivery and

Qualifications

• Six or more years in applied ML/geospatial modeling or forecasting.
• Strong Python and modern ML framework experience for production models.
• Experience with probabilistic evaluation, calibration, or uncertainty analysis.
• Experience with geospatial data, coordinate systems, and large spatial datasets.

Responsibilities

• Design, train, and deploy forecasting models across parcels, buildings, and regions.
• Build historical datasets with vintages, crosswalks, and reproducible features.
• Establish benchmarks across places, horizons, baselines, and public models.
• Measure accuracy, probabilistic scores, calibration, and uncertainty.
• Develop explanations and model cards for defensible practice.
• Own experiment tracking, data quality checks, and reproducible training pipelines.
• Collaborate with product engineers to deploy models with monitoring and rollback.
• Engage with public-sector reviewers to improve datasets and model behavior.
• Communicate methods and limitations in technical documents and materials.

Skills

Python

ML frameworks

Time-series evaluation

Geospatial data

Tools

PyTorch

GeoPandas

PostGIS

Job description

Worldcastr is seeking a founding Geospatial Machine Learning Engineer to turn our predictive spatial models into production-ready, validated forecasting capabilities. You will own end-to-end modeling across parcels, buildings, and regional indicators, building robust datasets and reproducible feature pipelines.

The role emphasizes probabilistic forecasting, geographic transfer, explainability, and rigorous evaluation across horizons, with a strong emphasis on production delivery and