Data Scientist – Program Evaluation & Analytics
- Compensation
- Not disclosed
- Location
- Tysons, VA, US
- Work setup
- Not specified
- Seniority
- Not specified
- Employment
- Contract
- Experience
- 6+ years
- Category
- Data science
- Posted
Skills
Must have
- Education / Training: Bachelor’s degree in statistics, biostatistics, epidemiology, data science, computer science, information systems, quantitative social science, econometrics, or a related field;
- Experience: 6+ years of progressively responsible experience spanning data analytics, program evaluation, business intelligence, application/data development, or a related quantitative discipline, wit
- SQL & Relational Data: Strong SQL skills and demonstrated experience working with complex relational data models, cohort construction, joins, data validation, and analysis-ready datasets.
- Python or R: Strong Python or R capability for reproducible analysis, including data manipulation, visualization, statistical modeling, and reporting; Python is preferred for the future-state analytic
- Analytical Judgment: Demonstrated ability to distinguish descriptive, predictive, and causal questions; select and defend appropriate methods; assess uncertainty and practical importance; and avoid co
- Reproducibility & Quality: Experience using version control, modular code, testing, peer review, documentation, and other practices that make analyses rerunnable when data change.
- Communication: Excellent written and verbal communication skills, including the ability to explain analytical findings and technical tradeoffs in plain language and work directly with stakeholders.
- Federal Data Environment: Sound judgment when working with sensitive healthcare, outcomes, public health, or program data, including PII/PHI and approved AI tools.
- Clearance: Ability to obtain a Public Trust clearance.
Nice to have
- Domain: Experience with VA, suicide prevention, mental health, healthcare outcomes, public health, or program-evaluation data, including an understanding of veteran referral, enrollment, treatment, di
- Advanced Methods: Experience with mortality, survival, longitudinal outcomes, propensity methods, weighting, matching, causal inference, or other cohort-balancing approaches.
- Modernization: Experience modernizing legacy analytical pipelines, especially migration from SAS to Python, with emphasis on reconciliation, testing, and maintainability rather than mechanical code tr
- Power Platform / BI: Hands-on experience with Power BI and Power Apps; ability to translate business requirements into usable dashboards or applications. Familiarity with Power Query, M, and DAX is pr
- Client Engagement: Experience presenting directly to non-technical clients and converting analytical results into program decisions or operational recommendations.
- Product / Delivery: Experience balancing hands-on technical contribution with delivery leadership, backlog or work prioritization, requirements gathering, and cross-functional coordination.
- Certifications: Microsoft Power Platform, Agile, project management, or related certifications are a plus.
Full description
Overview
LMI is seeking a mid-to-senior level Data Scientist to support the U.S. Department of Veterans Affairs (VA). This multifaceted role combines hands-on data engineering and visualization with applied program evaluation and statistical analysis. The position will translate complex program questions into validated datasets, reproducible analyses, and decision-ready dashboards and findings, using SQL and Microsoft Power Platform on the delivery side and Python or R for advanced analytical work.
The Data Scientist will serve as a bridge between program leadership, technical delivery, and analytical evaluation. The successful candidate will understand the underlying relational data and veteran journey, build and validate analysis-ready cohorts, develop Power Apps/Power BI solutions, independently select appropriate analytical methods, and communicate results in clear program language.
LMI is a new breed of digital solutions provider dedicated to accelerating government impact with innovation and speed. Investing in technology and prototypes ahead of need, LMI brings commercial-grade platforms and mission-ready AI to federal agencies at commercial speed.
Leveraging our mission-ready technology and solutions, proven expertise in federal deployment, and strategic relationships, we enhance outcomes for the government, efficiently and effectively. With a focus on agility and collaboration, LMI serves the defense, space, healthcare, and energy sectors—helping agencies navigate complexity and outpace change. Headquartered in Tysons, Virginia, LMI is committed to delivering impactful results that strengthen missions and drive lasting value.
Responsibilities
- Technical & Data Leadership: Lead the design, development, and delivery of analytics products that combine SQL-backed data, Microsoft Power Apps/Power BI, and reproducible analytical workflows.
- Program & Stakeholder Partnership: Partner with program leadership and business stakeholders to frame questions, clarify populations, outcomes, time windows, intended decisions, and analytical objectives; translate ambiguous needs into actionable technical and analytical requirements.
- SQL & Data Engineering: Write readable, efficient SQL against complex relational datasets; construct analysis-ready cohorts with explicit inclusion, exclusion, and time-window logic; and validate joins, grain, keys, duplicates, missingness, and denominators.
- Dashboards & Power Platform: Develop and maintain Power BI dashboards and Power Apps experiences that make program performance, outcomes, and operational information accessible and actionable; support data refreshes, usability, and maintainability.
- Applied Program Evaluation & Statistics: Independently select, apply, and defend appropriate analytical methods across descriptive, predictive, and causal questions, drawing on regression, longitudinal analysis, survival/time-to-event methods, cohort balancing, missing-data approaches, and model diagnostics as appropriate.