Graph Data Scientist

Graph Data Scientist

Hourly pay range:

$67.14-$69.33
Employment type:

Full-time
Schedule:

Approximately 1,920 hours annually
Location:

Primarily remote, with occasional onsite work in Washington, DC

Position Summary

The Graph Data Scientist will design and develop graph-based fraud-detection, entity-resolution, network-analysis, and knowledge-graph solutions using Neo4j or comparable graph technologies.

The position will analyze relationships among individuals, businesses, addresses, bank accounts, transactions, devices, awards, claims, applications, and other entities to identify organized fraud activity, hidden relationships, shared infrastructure, suspicious networks, and other indicators of fraud, waste, abuse, or mismanagement.

Primary Responsibilities
Design, develop, test, and maintain graph models supporting fraud detection, investigative analysis, entity resolution, and network discovery.
Develop and optimize Cypher queries or comparable graph-query logic.
Create graph schemas, node and relationship structures, indexes, constraints, and data models.
Apply graph algorithms such as centrality, community detection, shortest path, similarity, clustering, and link prediction.
Identify organized fraud rings, shared identifiers, hidden ownership, common addresses, common devices, coordinated transactions, and related risk patterns.
Develop graph-based features for machine-learning and fraud-risk models.
Apply statistical and machine-learning methods to graph-structured data.
Build and maintain knowledge graphs that integrate information from multiple public and non-public sources.
Support entity resolution, record linkage, identity matching, and relationship analysis.
Develop scalable graph ingestion, transformation, enrichment, and quality-control processes.
Evaluate graph-model performance, query performance, data quality, scalability, and investigative usefulness.
Develop reusable Python code, notebooks, graph algorithms, scripts, and technical documentation.
Produce network visualizations, link-analysis diagrams, graph-based findings, and investigative-support products.
Collaborate with investigative analysts to validate relationships and develop actionable leads.
Collaborate with forensic accountants to map transactions, ownership structures, and the movement of funds.
Collaborate with data engineers to ingest, normalize, and maintain graph-ready datasets.
Support the deployment, monitoring, maintenance, and improvement of graph analytics in production environments.
Brief government stakeholders on graph methodologies, findings, assumptions, limitations, and investigative implications.
Maintain documentation describing graph schemas, queries, algorithms, data sources, and analytic results.
Required Qualifications

Minimum of three years of hands-on experience using Neo4j or a comparable graph database.
Fluency in Cypher or a comparable graph-query language.
Minimum of three years of experience applying graph techniques to fraud detection or knowledge-graph solutions.
Strong understanding of graph theory, network topology, centrality measures, community detection, clustering, and shortest-path methods.
Minimum of three years of experience applying statistical or machine-learning techniques to graph-structured data.
Experience with clustering, classification, anomaly detection, graph features, or relationship-based risk modeling.
Strong Python programming skills.
Experience using commonly adopted data-science and machine-learning libraries.
Experience integrating and analyzing structured and unstructured data from multiple sources.
Experience designing or optimizing graph schemas, graph data pipelines, and graph queries.
Ability to create clear and accurate graph visualizations and link-analysis products.
Ability to explain graph methodologies and findings to technical, investigative, and executive audiences.
Strong analytical, documentation, communication, and presentation skills.
Ability to complete federal suitability, HSPD-12/PIV credentialing, and system-access requirements.
Preferred Qualifications

Experience applying graph methods to federal-benefit, financial-crime, public-integrity, or law-enforcement data.
Experience with Neo4j Graph Data Science, Databricks, Spark, SQL Server, Power BI, or comparable platforms.
Experience developing graph-based fraud indicators, knowledge graphs, or investigative lead-generation systems.
Experience with entity resolution, identity analytics, transaction networks, or organized-fraud detection.
Experience with i2 Analyst's Notebook or comparable link-analysis and visualization tools.
Degree in data science, computer science, statistics, mathematics, engineering, network science, or a related discipline.