Sanjay Kumar

Leading the Justice Tech Modernization Program by delivering impactful AI/ML solutions and cloud-based data platforms. Own the full product lifecycle from ideation to delivery, aligning business goals with technical feasibility to drive operational transformation across public safety and legal domains.

• AI Product Lifecycle Ownership: Led product lifecycle from concept to deployment for multiple AI/ML initiatives. Defined product strategy, success metrics, and technical feasibility in collaboration with engineering, data science, and public sector stakeholders.
• Generative AI Delivery: Built and launched a Legal AI Assistant using Azure OpenAI (RAG + fine-tuning) to support contract clause extraction, precedent search, and risk detection—cutting legal review time and improving compliance.
• Product Requirements & Roadmaps: Translated customer, legal, and operational needs into actionable product roadmaps and technical specifications. Balanced short-term priorities with a long-term vision to maximize impact across departments.
• Cross-Functional Development: Drove end-to-end product delivery with data science and engineering teams. Delivered high-quality AI features including VIN Decoding Product: Automated vehicle detail identification from officer-submitted data and Stay Away Order Detection Tool: NLP model to flag judicial mandates in court minutes, mitigating risk.
• Cloud Platform Enablement: Product-managed the implementation of a Medallion-based Azure Data Lake using ADF and Databricks, enabling scalable data ingestion, ML workflows, and secure analytics across agencies.
• User-Centered Insights & Analytics: Delivered intuitive Power BI dashboards, including repeat gun violence analytics using EPR and CAD data. Empowered agencies to take data-driven action based on near real-time insights.
• Data Governance & Automation: Championed data profiling, source-to-target mapping, and lineage tracking to improve quality. Co-led automation of data-cleaning workflows, reducing manual effort by 60% and improving trust in AI outputs.
• Team & Stakeholder Leadership: Managed a cross-functional team (2 Data Engineers, 1 Data Scientist). Facilitated alignment with legal, policy, and IT units to ensure timely delivery and successful adoption of AI solutions.
• Operational Excellence: Delivered scalable, cost-efficient ML pipelines and integrated DevOps and CI/CD workflows to accelerate time-to-market and increase system reliability.

Spearheaded the development and delivery of AI-powered sales optimization products, aligning machine learning initiatives with enterprise business goals. Delivered end-to-end ML solutions at scale, from ideation through deployment, while ensuring user impact, operational efficiency, and responsible AI adoption.

• Defined product vision and roadmap for AI-driven lead scoring solutions, aligning with sales strategy and improving win-rate predictions through predictive modeling.
• Led cross-functional teams (data science, engineering, DevOps) to develop scalable ML pipelines using Azure Databricks and ADF, integrated with Salesforce CRM data for real-time insights.
• Shipped AI features with full MLOps lifecycle support using Mflow, Docker, and Azure Kubernetes Service (AKS), ensuring reliability and scalability in production.
• Partnered with business stakeholders to translate requirements into data-driven products, driving adoption through intuitive dashboards and analytics via Tableau, Einstein Analytics, and TCRM.
• Oversaw data strategy, including migration from SQL Server to Azure Delta Lake, and designed data models and ETL workflows aligned to product outcomes.
• Leveraged ML techniques such as XGBoost, Random Forest, NLP, and time series forecasting (ARIMA/SARIMA) to power product features with measurable impact

Build and deploy machine learning models to improve transaction authentication rates and user experience by accurately predicting issuer declines in real-time payment processing. Partnered with business stakeholders, engineering, data science, merchant tech, and compliance teams to translate complex AI concepts into platform-ready products delivering measurable business value.

• Led data preprocessing and exploratory data analysis (EDA), including outlier detection, missing value treatment, feature scaling, and domain-specific feature engineering to enhance signal quality for modeling.
• Evaluated and fine-tuned multiple ML models (Neural Networks, GBM, Random Forest) using performance metrics such as F1-Score, ROC-AUC, and A/B Testing. Deployed GBM model due to its optimal trade-off between precision and recall.
• Used Simility, an in-house ML platform, to deploy and operationalize models in a scalable and monitored environment, integrated with fraud analytics workflows.
• Explored fraud trends using Tigress, an internal data aggregation and visualization tool, to uncover patterns and refine model inputs.
• Built and maintained models on Vertex AI, leveraging transactional data stored in Google BigQuery; used cross-validation and batch testing for robust generalization across geographies and customer segments.
• Wrote advanced SQL queries (Hive, Teradata) for data extraction and feature generation, including window functions and subqueries for longitudinal analysis.
• Created Tableau dashboards to communicate model insights and fraud trends to risk analysts and stakeholders.
• Maintained experiment tracking, documentation, and agile workflows using Jira and Confluence.

Led the development of machine learning models to predict 30-day readmission risk in heart failure patients using EMR data, resulting in 2% improvement in case identification

• Developed end-to-end ML pipelines: data extraction, profiling, feature engineering, model training (Logistic Regression, GBM, Neural Nets), validation, and deployment.
• Integrated Social Determinants of Health (SDoH) from Claritas Prism with membership and provider data via Talend ETL into MS SQL Server.
• Performed advanced EDA using Python (Pandas, Matplotlib, Seaborn) to identify care gaps and efficiency improvements.
• Applied SHAP and LIME for model interpretability to support provider decision-making.
• Engineered features using ICD-10/CPT codes, HEDIS, STAR metrics, MARA, and MTM data for care and claims analytics.
• Built an opioid overdose prediction model from 30M+ pharmaceutical claims records, enhancing early risk detection.
• Wrote complex SQL queries and optimized stored procedures in MS SQL Server and Azure, redesigned data models for performance gains.
• Delivered insights via Tableau dashboards and Python visualizations to business stakeholders and clinical teams.
• Applied statistical methods (T-tests, ANOVA, PCA, Time Series) to validate trends and experiment results.
• Contributed to A/B testing strategies for improving member experience and healthcare site execution.

Built scalable People Analytics solutions to optimize talent acquisition, retention, and workforce planning, aligning data strategy with HR and business goals.

• Collaborated with cross-functional teams to gather requirements and develop automated analytics workflows using Alteryx, integrating data from Workday, SAP BPC, Salesforce, SAP Fiori, and Oracle Primavera P6, and publishing consolidated HANA DataMarts.
• Designed and deployed interactive Tableau dashboards to visualize hiring trends, attrition, labor utilization, time tracking, and open requisitions across departments and geographies.
• Conducted diversity analysis across age, gender, ethnicity, and reporting structures to uncover insights on inclusion, engagement, and enablement.
• Developed predictive models using Auto-Regressive and Survival Analysis techniques for headcount, attrition, and transfer forecasting.
• Applied Python (Pandas, NumPy, Matplotlib, Seaborn) for data cleaning, analysis, and visualization; engineered complex SQL queries and stored procedures for data extraction.
• Deployed BI solutions via Tableau Server, with version control and CI/CD support through GitLab and Jenkins.

• Designed and implemented end-to-end BI solutions using SSRS, SSIS, SQL, Python, and R to support enterprise analytics and reporting. Developed dynamic SSRS reports, created SSIS packages for ETL processes, and performed statistical analysis and visualization using Python and R. Wrote optimized SQL queries, including complex Joins, CTEs, stored procedures, and triggers to extract and transform data for actionable insights.