Graph Data Scientist

The PRAC analytic platform operates in a cloud-based environment using tools such as Azure Databricks, Microsoft SQL Server, Microsoft Power BI, and neo4j, along with other advanced technologies to analyze and visualize public, non-public, and commercially available data. The Graph Data Scientist will work within agile sprints and Integrated Project Teams (IPTs), collaborating with PRAC stakeholders, OIG partners, and law enforcement to deliver high-quality analytic outputs that support oversight, investigative support, and continuous capability enhancement. The role requires strong technical expertise, analytic rigor, and the ability to translate complex graph-based insights into clear, actionable intelligence for diverse stakeholders., * Develop graph-based analytic rules and models to detect fraud, waste, abuse, and mismanagement across multi-agency and multi-program datasets.

• Apply advanced graph theory concepts-including network topology, centrality measures, community detection, and shortest-path algorithms-to identify hidden relationships and emerging fraud schemes.
• Use Neo4j (or similar graph databases) and fluently write Cypher queries to build, test, and optimize graph models and knowledge graphs supporting PRAC analytic use cases.
• Conduct exploratory data analysis and data profiling on public and non-public datasets to prepare graph structures and ensure high-quality ingestion.
• Apply clustering, classification, anomaly detection, and other machine learning techniques to graph-structured data to enhance fraud detection capabilities.
• Design, implement, and optimize graph data pipelines, schemas, and data models that support high-complexity networks and large-scale analytic environments.
• Document methodologies, graph models, code, and analytic outputs in accordance with PRAC data governance, EDM standards, and federal oversight requirements.
• Collaborate with PRAC stakeholders, OIG partners, and law enforcement to refine analytic approaches, validate graph-based findings, and support cross-program risk identification.
• Conduct graph-based investigative analysis using open-source, law enforcement, and third-party data to support lead generation and case development.
• Build link analysis diagrams, entity relationship charts, and graph-based intelligence products that visualize complex networks and relationships across multiple data sources.
• Use graph analytics to validate PRAC-developed analytic rules and model outputs, performing due diligence to confirm or refute potential fraud indicators.
• Support OIG and law enforcement partners by producing customized investigative work products and presenting graph-derived insights in clear, actionable formats.
• Follow PRAC investigative workflows, templates, and policies for handling law-enforcement-sensitive data and multi-agency investigative information.
• Evaluate emerging graph analytics tools, graph databases, and data sources; provide recommendations aligned with PRAC mission needs and federal policy.
• Collaborate with PRAC’s OCIO, OCISO, and OCDO to configure, integrate, and optimize approved graph tools within the PRAC analytic environment.
• Support ingestion of new data sources into graph environments, including schema design, transformation logic, and quality control.
• Assist PRAC in analyzing cutting-edge graph-based AI/ML tools and technologies, providing technical documentation and risk considerations to support Government procurement decisions.
• Maintain awareness of evolving fraud schemes, graph analytic techniques, and cross-program risks to inform future capability development.
• Work closely with data scientists, investigators, statisticians, and engineers to integrate graph analytics into broader fraud detection and investigative workflows.
• Communicate complex graph-based findings to senior PRAC leadership, OIG partners, and non-technical stakeholders in clear, accessible terms.
• Ensure analytic integrity, reproducibility, and compliance across all graph-related workstreams.
• Contribute to continuous improvement of analytic processes, data quality, and graph-based fraud detection methodologies.

• Three (3) or more years of hands-on experience using Neo4j or similar graph databases, including fluency in Cypher or comparable graph query languages.
• Deep understanding of graph theory and network science, including network topology, centrality measures, community detection, and shortest-path algorithms.
• Three (3) or more years of hands-on experience applying statistical and machine learning techniques-such as clustering, classifiers, and anomaly detection-to graph-structured data.
• Three (3) or more years of experience applying graph methods to fraud detection, knowledge graph development, and identification of hidden relationships.
• Experience designing, implementing, and optimizing graph data pipelines, graph data models, and schemas supporting high-complexity networks.
• Strong Python programming skills, including use of standard machine learning libraries (e.g., Pandas, NumPy, Scikit-Learn, NetworkX).
• Strong oral and written communication skills., * Experience supporting fraud analytics, investigative support, or program integrity efforts within federal oversight environments (e.g., OIGs, PRAC, CIGIE, law enforcement).
• Experience integrating multi-agency or multiprogram datasets into graph environments to detect cross-program risks and complex fraud schemes.
• Experience producing graph-based investigative work products such as link analysis diagrams, entity relationship charts, and network visualizations.
• Experience developing or deploying AI/ML models applied to graph data.
• Experience supporting enterprise data management (EDM), data governance, and federal data quality standards.
• Familiarity with PRAC AI governance policies, OMB memoranda (M-25-22, M-25-21, M-26-04), and the NIST AI Risk Management Framework.
• Advanced certifications in data science, graph analytics, or fraud examination (e.g., CFE, Neo4j Certified Professional).

Required Education: Bachelor’s degree (BA/BS) in Data Science, Computer Science, Mathematics, Statistics, Engineering, Information Systems, or a related quantitative field.

Preferred Education: Master’s degree (MS/MA) in Data Science, Analytics, Computer Science, Statistics, or a related discipline.

• Health (PPO & HDHP) Insurance, Dental, Vision, STD & LTD, Basic Life Insurance, 401k Company Match, & Voluntary Products.

About Knowesis Inc.

201-500