Over a decade shipping AI, ML & geospatial products at scale.
From NASA JPL's Methane Campaigns to Apple Maps Launches
Twelve-plus years leading 0-to-1 initiatives across Apple, NASA JPL, and public & private sector teams — with deep expertise in computer vision, spatial data intelligence, and applied ML on scalable systems.
Shaped cross-functional roadmap and launch strategy for the navigation substrate and geospatial data layers powering Apple Maps' international rollout. Championed attribution and codebase upgrades for 4M+ special road and lane features feeding the routing algorithm.
Drove delivery of ML models (Random Forest, PCA, UMAP) across 41 behavioral and sensor signals — classifying and updating 100k+ events monthly across 40+ countries.
Developed & published the Vista-CA GSAAM — a geospatial ML algorithm for the California Air Resources Board that tracks 1M+ energy, waste, and agriculture features at >99% accuracy.
Collaborated on CNN models applied to hyperspectral imagery from NASA JPL's AVIRIS-NG spectrometer to detect and quantify methane plumes at continental scale. Published in Nature. Now powering carbonmapper.org.
Setting long-horizon roadmaps that translate research and market signal into shipped systems.
Computer vision, classification, spatial modeling — from prototype to production pipeline.
Location intelligence, spatial data ingestion, and mapping systems at continental scale.
Aligning senior leadership on bets, tradeoffs, and multi-year investments.
Building and structuring product teams for durable throughput.
Programmatic ingestion, pipeline design, and analytics that unlock decision-making.
Engineering, science, design, and business — driving alignment across disciplines.
Building for both power users at scale and the everyday person opening the app.
Deep GIS foundations — the discipline underneath a decade of spatial products.
Open to conversations about product leadership, AI/ML platforms, and geospatial systems at scale.
talha.rafiq.phd@gmail.com →