Senior Data Scientist / AI-ML & Anomaly Detection Lead

Accountabilities:

• Lead the modernization of an existing 11-rule proof-of-concept library, evolving deterministic analytics into a scalable enterprise detection framework.
• Architect transaction-focused anomaly detection approaches, defining relevant data features, risk scores, tolerance thresholds, alert priorities, and detection criteria.
• Develop, refine, and deploy supervised and unsupervised machine learning models where they provide measurable value beyond conventional business rules.
• Apply advanced statistical and machine learning techniques including classification, clustering, forecasting, feature engineering, and anomaly detection.
• Establish rigorous model evaluation and calibration processes, including accuracy measurement, threshold optimization, and detailed false-positive and false-negative analysis.
• Ensure all analytical models remain explainable and transparent, with human-readable reasoning and supporting evidence for every material anomaly or flagged transaction.
• Build audit-ready validation datasets and technical test processes capable of demonstrating model accuracy, reproducibility, and compliance with defined performance thresholds.
• Provide technical documentation, configuration baselines, test scripts, and system logs that enable independent government validation teams to reproduce and rerun analytical routines.
• Implement monitoring approaches for model scoring, feature mapping, version control, data drift, and ongoing analytical performance.
• Develop standardized methodologies for measuring the value generated by analytical insights, including potential cost avoidance, financial recovery, risk exposure, and investigative referrals.
• Work with complex financial, procurement, contract, purchasing-card, and other transaction-level datasets to uncover patterns and potential risks.
• Maintain a strong focus on scientific reproducibility, technical traceability, and defensible analytical evidence throughout the project lifecycle.
• Support technical validation, milestone acceptance, and performance tuning activities as required, including periodic travel to Tallahassee, Florida.

Requirements:

• 10+ years of comprehensive experience in data analytics and data science, including at least 5 years of hands-on machine learning engineering experience.
• Demonstrated experience delivering data science, predictive modeling, or advanced analytics solutions within federal, state, military, or local government environments.
• Advanced hands-on proficiency with Python and SQL for complex data manipulation, analysis, engineering, and model development.
• Deep expertise in supervised and unsupervised machine learning, classification, clustering, statistical forecasting, anomaly detection, and advanced feature engineering.
• Proven ability to evaluate and calibrate models, establish appropriate thresholds, and analyze the operational impact of false positives and false negatives.
• Experience translating model outputs into defensible, audit-ready evidence rather than relying solely on probability scores or opaque predictions.
• Strong understanding of scientific reproducibility, including the creation of technical documentation, configuration baselines, validation artifacts, and materials that support independent third-party replication.
• Experience working with highly diverse, transaction-level datasets, particularly financial, procurement, contract management, purchasing-card, or related ledger data.
• Prior experience developing fraud, waste, abuse, improper-payment, financial-crime, or corporate-risk analytics is strongly preferred.
• Experience working with oversight-oriented data from government agencies or comparable public-sector environments is a plus.
• Hands-on experience with Azure Databricks, MLflow, or Azure Machine Learning in secure government cloud environments is preferred.
• Familiarity with hybrid detection architectures combining deterministic business rules with machine learning-based anomaly detection is highly desirable.
• Experience with model cards, explainability frameworks, and structured human-in-the-loop validation workflows is a strong advantage.
• Must be a U.S.-based citizen or resident and able to successfully complete an FDLE Level II background screening, including fingerprinting, within 5 business days of contract award.
• Candidates should be prepared to demonstrate specific government project experience, including the agency served, analytical problem addressed, technical approach, individual contribution, deployment and validation status, and measurable outcomes.

Benefits:

• Part-time consulting engagement averaging 14–18 hours per week.
• Remote work within the United States.
• Flexible workload structure, with additional hours concentrated around major project milestones, technical validation, and government acceptance activities.
• Milestone- and deliverable-based fixed-price compensation structure.
• Opportunity to contribute to a high-visibility government Decision Intelligence initiative addressing financial oversight, risk detection, and fraud, waste, and abuse.
• Exposure to advanced machine learning, anomaly detection, data engineering, and explainable AI applications in a complex public-sector environment.
• Periodic travel to Tallahassee, Florida, as required by project activities.
• Opportunity to work on a technically demanding initiative involving large-scale, multi-agency data and measurable analytical performance requirements.