ML Engineer Security Clearance

Position: ML Engineer with Security Clearance
Full Description Are you passionate about designing AI systems that unlock the full power of geospatial intelligence? At Gateway Geospatial Group (G3), we  building the future of GEOINT by turning fragmented data into a semantically searchable, AI-enriched ecosystem. Responsibilities Design and implement NLP and semantic enrichment pipelines that drive next-generation data discovery.

• Develop, fine-tune, and deploy transformer-based models (e.g., BERT, Sentence Transformers) to generate vector embeddings and metadata summaries.

• Build machine-learning-based methods for semantic search, automated tagging, and similarity matching across multi-modal GEOINT datasets.

• Work alongside Dev Ops, data engineers, and intelligence analysts in an Agile environment to deliver AI-ready infrastructure.

• Analyze temporal, geographic, and mission-specific patterns to build graph relationships across disconnected data sources.

• Contribute to real-world solutions supporting the National Geospatial-Intelligence Agency (NGA) and broader defense intelligence missions.

Qualifications:

TS/SCI clearance (required).

• Experience building or fine-tuning NLP models, especially for semantic understanding or classification.

• Strong understanding of vector similarity search techniques (e.g., FAISS, Annoy).

• Hands-on experience with Python, PyTorch, Hugging Face Transformers, scikit-learn, and Apache Airflow.

• Familiarity with graph-based data models (e.g., Neo4j) and knowledge graphs.

• Ability to create clean, auditable machine learning pipelines that integrate with structured ETL workflows.

• Comfort working across multiple geospatial formats (GeoJSON, WKT/WKB) and standards (ISO, OGC).

• A strong grasp of security-first AI development, especially in IC or DoD environments. Desired Qualifications Prior experience working on government or IC contracts involving GEOINT data.

• Familiarity with Graph

QL, PostGIS, and/or Elastic Search in AI-powered systems.

• Experience designing AI solutions that comply with ICD 503 and zero-trust architecture principles.

• Knowledge of multi-modal data enrichment and human-in-the-loop training workflows.