Imagery Scientist; SAR

Geo Owl
· SDVOSB
· Est. 2013

NGA MAVEN PROGRAM
· SYNTHETIC APERTURE RADAR
· GEOINT
· EXPERT
· KEY POSITION

The Imagery Scientist is the subject matter expert on synthetic aperture radar. You will provide technical direction and conduct the work necessary to acquire and prepare SAR imagery that meets the Government’s specified quality, standards, and requirements. You will integrate emerging SAR sensors into Maven pipelines, develop pre‑processing and tiling workflows, generate pre‑labels, and provide geospatial and imagery expertise to improve data curation for machine‑learning algorithm training and evaluation.

• Assess emerging SAR sensors: evaluate differences in collection characteristics, metadata, data formats, and phenomenology compared to currently utilized platforms and recommend ETL adaptations for schema and API changes.
• Execute daily SAR imagery curation: assess image quality using RNIIRS and SAR‑specific metrics, prioritize acquisitions per government direction, and convert files to required formats.
• Tile full‑size SAR imagery to specified pixel dimensions or geospatial bounds; develop and implement mathematical conversion models to transform labels across imagery types (PNG to standard SAR imagery formats) and between orthorectified and non‑orthorectified imagery.
• Generate pre‑labels from intelligence reporting, machine‑derived observations, and human observations—conforming to defined ontology standards to guide human labelers to objects of interest.
• Identify and provide EO imagery coincident with SAR deliveries within specified geographic and temporal ranges; explore options for coincident collects across platforms.
• Oversee weekly deliveries of processed imagery and execute ad‑hoc quick‑turn deliveries as operational needs arise.

The SAR data you curate, tile, and pre‑label becomes training material for NGA Maven’s AI/ML models. The precision of your science—how well the imagery reflects the phenomenology the model needs to learn—directly determines how those models perform against real‑world targets. As a key position, your SAR expertise anchors Geo Owl’s technical depth on Maven and distinguishes our ability to deliver across both EO and SAR modalities.

• Integrate emerging SAR sensors and platforms into existing Maven data pipelines; assess metadata, format, and schema differences and adapt ETL processes accordingly.
• Develop and implement mathematical conversion models to transform data labels across imagery types (PNG to standard SAR imagery formats) and between orthorectified and non‑orthorectified imagery.
• Execute tiling and pre‑processing of full‑size raw SAR imagery to specified formats and geospatial dimensions while maintaining metadata integrity and data quality standards.
• Analyze and assess SAR image quality and sensor metadata to curate acquisitions in alignment with government‑directed priorities.
• Generate pre‑labels from intelligence reporting, machine‑ and human‑derived observations; provide coincident EO imagery within specified geographic and temporal ranges.
• Identify and implement curation methods—including NLP‑based intelligence extraction and automated machine techniques—to maximize high‑value imagery yield.
• Monitor and prioritize data holdings to support data diversity needs (geographic regions, temporal windows, metadata and scene characteristics, sensor platforms).
• Develop, test, and evaluate new SAR algorithms, methodologies, and products using advanced processing tools and cloud‑based solutions.

• Active TS/SCI clearance.
• Minimum 18 experience points required.
• 4+ years as a SAR expert with deep understanding of collection, phenomenology, image‑formation process, and exploitation products.
• Experience using SARPy and MATLAB SAR toolbox.
• Experience with RNIIRS, information‑theoretic image‑quality metrics, SAR imagery quality metrics (Integrated Sidelobe Ratio, Multiplicative Noise Ratio), and sensor metadata describing geometry impacts on phenomenology (graze, squint, azimuth).
• Experience exploiting SAR to determine the occurrence and location of objects of…