Curve & Compass Analytics provides rigorous program evaluation and applied data science services for government agencies, universities, and nonprofit organizations that need credible evidence about whether their programs are working — and analysis that holds up to funder and oversight scrutiny.
What We Do
Stakeholders, funders, and oversight bodies increasingly expect more than pre/post comparisons and narrative summaries. They expect causal rigor, transparent methods, and findings that can withstand a technical review. Curve & Compass Analytics brings quantitative depth and applied machine learning to evaluation questions that traditional approaches handle poorly — producing evidence that is methodologically sound, clearly communicated, and actionable.
Who We Serve
— State and local government agencies administering federally funded programs
— Universities and research institutions conducting sponsored evaluations
— Nonprofit organizations demonstrating program impact to funders
— Federal agencies and contractors requiring independent evaluation support
— Organizations with grant requirements for third-party evaluation
What We Deliver
Impact Evaluation and Causal Inference We design and execute evaluations that go beyond correlation to establish whether programs produce the outcomes they intend. Our methods include difference-in-differences analysis, synthetic control methods, regression discontinuity designs, and causal forest approaches — answering not just whether a program worked, but for whom and under what conditions.
AI and Algorithmic Program Evaluation As government agencies increasingly deploy AI in eligibility determination, risk scoring, and resource allocation, there is a growing need to evaluate whether those systems are functioning as intended and producing equitable outcomes. We bring technical AI/ML expertise to evaluation questions that traditional program evaluators are not equipped to answer.
Process Evaluation and Implementation Analysis Understanding why a program produced the outcomes it did requires examining how it was actually implemented. We conduct systematic implementation analyses that identify fidelity gaps, contextual moderators, and operational factors that explain outcome variation across sites, cohorts, or time periods.
Mixed-Methods Evaluation Design We design evaluations that integrate quantitative and qualitative methods — using statistical findings to sharpen qualitative inquiry and qualitative evidence to interpret quantitative results — producing a more complete picture than either approach alone.
Predictive Modeling and Applied Data Science Beyond evaluation, we provide applied data science services for organizations ready to move from descriptive reporting to predictive capability — forecasting program demand, identifying at-risk populations, modeling resource allocation scenarios, and building explainable models that program staff can use and trust.
Data Infrastructure and Reporting Assessment Many evaluation limitations are data infrastructure limitations in disguise. We assess existing data collection, management, and reporting systems — identifying gaps that constrain evaluation capacity and recommending practical improvements that strengthen the evidence base for future work.
Our Perspective
The most common failure mode in program evaluation is the gap between methodological ambition and analytical execution — designs that promise causal inference but deliver descriptive statistics, or machine learning applied without interpretability. We build evaluations the way we build everything else at Curve & Compass: designed to withstand scrutiny from the start.
Engagement Process
Initial consultation (no charge) — we review your evaluation needs and data environment
Scope development — evaluation design, timeline, and deliverable specification
Engagement agreement — standard terms
Data collection, analysis, and reporting
Ongoing technical assistance available for multi-year evaluations
Project pricing and hourly rates provided upon inquiry.