Graph Data Scientist
Key Personnel
Role Title | Graph Data Scientist |
Reports To | Program Manager / Lead Data Scientist |
Type | Key Personnel — full-time for the period of performance |
Primary Outcome | Design and operate graph-based analytics that expose fraud rings, synthetic identities, and hidden relationships across large-scale federal benefit data — turning interconnected data into actionable investigative leads. |
Role Summary
We are seeking a Graph Data Scientist to design, build, and operate graph-based analytic solutions that detect fraud, expose hidden relationships, and strengthen program integrity across large-scale federal benefit data. This is a hands-on technical role for a specialist who sees relationships where others see disconnected data — someone fluent in graph theory, Neo4j, and graph-based machine learning who can model complex networks, engineer scalable graph pipelines, and surface the non-obvious connections that traditional analytics cannot detect. The person in this role applies graph algorithms, statistical methods, and machine learning to uncover fraud rings, synthetic identities, and anomalous network behavior, and translates those findings into investigative products that accelerate lead generation and case development.
The ideal candidate pairs deep graph-database and Cypher expertise with a strong machine-learning foundation, production-grade Python skills, and the judgment to build models that hold up at federal scale and under investigative scrutiny.
Key Responsibilities
• Design, develop, and maintain graph-based analytic solutions supporting fraud detection, investigative analysis, and program-integrity initiatives.
• Build and optimize graph databases, graph schemas, and knowledge graphs using Neo4j or comparable graph-database technologies.
• Develop Cypher (or comparable) graph queries to identify hidden relationships, fraud rings, suspicious networks, synthetic identities, and other complex entity relationships.
• Apply graph algorithms — centrality, community detection, shortest-path, similarity, and pathfinding — alongside statistical analysis and machine learning to identify emerging fraud patterns and anomalous network behavior.
• Apply clustering, classification, and anomaly-detection techniques to graph-structured data drawn from public and non-public sources.
• Design, implement, and optimize graph data pipelines, data models, and schemas that support large-scale, high-complexity networks.
• Develop interactive graph visualizations, relationship maps, and link-analysis products that accelerate lead generation, case development, and investigative decision-making.
• Collaborate with investigative analysts, forensic accountants, data engineers, and analytics leadership to deliver integrated analytic solutions.
• Document data models, methodologies, model performance, and validation results, and maintain accuracy and evidentiary integrity across all work products.
Required Qualifications
• Three (3) or more years of hands-on experience using Neo4j or a similar graph database, with fluency in Cypher or a comparable query language, to detect potential fraud using leading-edge techniques and best practices. (mandatory)
• Deep understanding of network topology, centrality measures, community detection, and shortest-path algorithms, applied across a multitude of public and non-public data sources. (mandatory)
• Three (3) or more years of hands-on experience in statistical and machine-learning foundations — clustering, classifiers, and anomaly detection — as applied to graph-structured data. (mandatory)
• Three (3) or more years of hands-on experience applying graph methods to fraud detection and knowledge graphs. (mandatory)
• Strong Python skills using standard machine-learning libraries. (mandatory)
• Experience designing, implementing, and optimizing graph data pipelines, data models, and schemas that support large-scale, high-complexity networks — preferably within large federal benefit programs.
Preferred / Differentiators
• Hands-on experience with named Python graph and ML frameworks such as NetworkX, Neo4j Graph Data Science (GDS), Pandas, Scikit-learn, and PyTorch Geometric (or comparable).
• Experience building Graph Neural Network (GNN) models (e.g., GraphSAGE) for fraud detection at scale, including handling new nodes without full retraining.
• Cypher query optimization experience — index utilization and query planning (EXPLAIN) — and use of the official Neo4j Python driver / GDS Python client.
• Neo4j Graph Data Science certification, or equivalent demonstrable graph-ML credentialing.
• Experience deploying graph analytics in cloud-native / Lakehouse environments (e.g., Azure Databricks, Microsoft Fabric, Azure Data Lake, SQL Server, Power BI, Git).
• Demonstrated work on synthetic-identity detection, entity resolution, or organized fraud-ring discovery.
• Prior experience supporting an Office of Inspector General (OIG), law enforcement, or a federal financial-crime / program-integrity mission.
Never miss an opportunity! Create an alert based on the job you applied for.