Our Insights
Dive deeper and explore our white papers, case studies, and presentations. Each project reflects our drive to deliver thoughtful, effective solutions.
Synthetic Data & Statistical Innovation
WHAT ARE SYNTHETIC DATA?
Synthetic data recreate the patterns and complexity of real-world information without relying on actual records. It gives teams a safer, faster way to test, develop, and innovate.
Watch: Synthetic Data: Innovating Without Risking Real Data | Real Reveal
Synthetic data are changing how organizations safely share and analyze sensitive information. Our research explores how these approaches can support secure innovation across federal environments.
WHITE PAPER
Synthetic Data Applications for Federal Statistics
This paper explores the diverse ways in which synthetic data are used in a federal context, the unique concerns federal agencies face when integrating it into an existing workflow and presents solutions that have worked to deploy synthetic data systems.
AI & Automation
From legacy system modernization to enhanced data workflows and accelerated development cycles, our solutions help you operate more efficiently while freeing teams to focus on strategic, high-impact initiatives rather than repetitive work.
WHITE PAPER
Ensemble Retrieval Strategies for an Improved NAICS Search Engine in the U.S. Census Bureau
Joint Statistical Meetings (JSM) 2024
This project redesigns the existing North American Industry Classification System’s (NAICS) string-matching search engine in the U.S. Census Bureau with a semantic search engine, achieving an accuracy of 86.87% on synthetic data, a 45.1% improvement over a keyword search.
PRESENTATION
Operationalizing the Hard-to-Count Framework: Integrated Methods for Improving Census Coverage
Joint Statistical Meetings (JSM) 2026
This research developed in collaboration with NORC at the University of Chicago on a question central to the future of federal statistics: How can organizations improve census coverage for hard-to-count populations?
WHITE PAPER
New Construction Activity Detection and Classification via Satellite Imagery and Machine Learning
AGX Applied Geography Conference 2025
This experimental data product transforms construction monitoring with satellite imagery and computer vision for faster decisions, lower risk, and improved operational efficiency, producing data that can be integrated into portions of the Census Bureau’s economic indicator series.
WHITE PAPER
Automating Census Legislative Research
Federal Committee on Statistical Methodology (FCSM) 2024
This paper explores the diverse ways in which synthetic data are used in a federal context, the unique concerns federal agencies face when integrating it into an existing workflow and presents solutions that have worked to deploy synthetic data systems.
Code Conversion & Translation
Across both multilingual communication and legacy system modernization, our approach emphasizes scalable AI augmentation that pairs machine efficiency with human expertise to deliver accurate, trustworthy, and federal-standard solutions.
WHITE PAPER
SAS to Python AI-Driven Translation
Federal Computer Assisted Survey Information Collection Workshops (FedCASIC) 2025
Joint Statistical Meetings (JSM) 2025
This paper explores how generative models can be used to engage with the SAS translation problem effectively, reducing administrative burden, improving code consistency, and transforming codebases into an open-source framework.
PRESENTATION
Exploring AI vs. Professional Translation: A Mixed-Methods Approach
American Association for Public Opinion Research (AAPOR) 2025
This presentation addresses the significant challenges the federal statistical system faces in collecting data from non-English speaking populations. We explore Large Language Models (LLMs)’s potential yet unproven solution for improving translation accuracy and efficiency.