Reveal data scientists Cameron Milne, Ellie Mamantov, and John Lynagh will present a new SAS to Python product at FedCASIC on April 23, 2025. The product is supporting the U.S. Census Bureau’s migration of SAS programs to Python by performing automated translations quickly and inexpensively compared with other code translation products on the market. Read more about Reveal’s work migrating SAS Codebases to Python here.

On March 11-12, 2025, the Reveal team had the privilege of attending the Intel Public Sector Summit in Washington, D.C., where industry leaders, government innovators, and technology experts came together to address critical developments in innovation, security, AI, and operational efficiency. The event provided an invaluable opportunity to engage with cutting-edge solutions, gain insights into emerging technologies, and discuss collaborative approaches to solving complex public challenges.

Key Takeaways

1. Partnerships are Necessary: Intel and Amazon are leading a partnership with the Census Bureau to change the way economic data collection happens away from a survey-centered paradigm to a direct company-to-government partnership. This new paradigm, that Reveal is also supporting, ensures faster and more reliable access to data. At Reveal, we are exploring similar partnerships and solutions, embedding computer vision and satellite imagery technologies to deliver greater value to other Federal agencies, commercial and SLED (State, Local, and Education) customers, driving innovation and operational efficiencies across these sectors.

2. Powering AI with CPUs: Intel’s new 18A chips could be game changers for advanced AI, working alongside GPUs to boost performance. For many use cases, existing servers powered by CPU’s are capable of running AI and ML models which increase efficiency and allow personnel to spend time on higher-valued tasks. Leveraging existing infrastructure will be essential for Government agencies to innovate while keeping costs down.

3. Small but Mighty: Small language models (SLMs) are the way to go for quick, cost-effective AI deployments that deliver immediate results with less risk. While real-time data processing and decision-making using Large Language Models (LLMs) within data centers is critical, deploying SLMs on the edge devices with limited hardware and processing capabilities can be highly beneficial. Tailored SLMs designed for specific analytical needs can significantly enhance edge computing, particularly in industries such as supply chain, energy, and transportation.

4. The Human Touch: Human-centered AI design will be essential for Government agencies to accomplish progress in the AI world since it requires close collaboration between teams and, often, a culture shift on data acquisition and data handling. When working with GenAI and LLMs, inference also plays a crucial role in shaping the outcomes. During a breakout session, partners from the Census Bureau and Intel discussed how AI efforts are as much about business as they are about technology. Human subject matter expertise and change management principles and processes coupled with GenAI can enhance accuracy, scalability, efficiency, and adoption in data transformation.

5. The Time is Now: The public sector needs to dive into new tech head-on to keep up with the fast-paced innovation of the private sector. It is through summits and events like this that all stakeholders can come together to collaborate and construct the future together.  

March 2025

Reveal is excited to support the Conference on Research in Income and Wealth (CRIW) held by the National Bureau of Economic Research (NBER) in Alexandria, VA. The conference is backed  by the U.S. Census Bureau and the U.S. Bureau of Labor Statistics through an award to Reveal. Centered around the Changing Nature of Work, the conference has influential speakers with both academic and federal backgrounds in economic research. The Conference was directly relevant to the work the Reveal Transformation Team did in support of Census as they transitioned to their new headquarters throughout 2024-25.

January 2025

Reveal is partnering with Rutgers University's SEED2S+X Program to provide experiential learning opportunities for their students through an Externship program. Reveal's Innovation Lab developed curriculum for Rutgers Students to apply their data science studies to federally scoped challenges. As Rutgers is a land grant university, students are tasked with using public United States Department of Agriculture (USDA) data sets to hone their technical capabilities. 

January 2025

In support of our federal partner Federal Motor Carrier Safety Administration (FMCSA), Subash Thota and Taylor Wilson attended the Transportation Research Board Annual Meeting with Vision Zero, a nonprofit campaign to reduce the number of motor carrier-related deaths to zero. This solution was developed and implemented by Reveal for the FMCSA.

December 2024

Reveal won the Department of Commerce Bronze Medal in for its work on StatVentures and Topx Toolkit. The award is given annually to an organization that demonstrates superior performance in impacting the DOC or one of its units' missions.

December 2024

Reveal earned a Level 3 maturity rating under the Capability Maturity Model Integration (CMMI) framework for Services and Development. The CMMI framework sets the standard for optimizing performance, ensuring product quality, and maintaining consistent delivery. Achieving a Level 3 rating reflects Reveal’s strong commitment to continuous process improvement at both the organizational and individual project levels. But this isn’t just about us—it’s about delivering more value to our clients. Our focus on refining processes means we’re always improving, guaranteeing better outcomes and real solutions to their greatest challenges.

August 2024

Senior Data Scientist Cameron Milne presented Ensemble Retrieval Strategies for an Improved NAICS Search Engine in the U.S. Census Bureau at the 2024 Joint Statistical Meetings (JSM) in Portland, Oregon. The presentation and later journal submission details his experiments in building a more performant search engine using Large Language Models (LLMs). Experiments include replacing sparse embeddings with dense embeddings, chunking documents and combining context windows, applying complementary re-ranking algorithms, and more. Dr. NAICS, the platform where the search engine is available, is among the first known implementations of AI in the U.S. Census Bureau.

Read the full story here and the submission paper here.

June 2024

Reveal partners with University of Michigan's School of Information on data science projects designed to expand learning opportunities for students interested in working in or with the Federal government. Projects range from assessing OCR model performance at the National Archives along with UI/UX support for self-serve applications at the U.S. Census Bureau. 

June 2024

Reveal attended the AWS Summit in Washington to learn how the Intel Corporation and Amazon Web Services are coming together to support Reveal and its federal partners to produce innovative data processing solutions within the U.S. Census Bureau. 

April 2024

Reveal hosted 25 Penn State students at its headquarters in Fulton, MD for an introductory discussion on data science in the federal space. Elo Lewis, Data Scientist; Taylor Wilson, Vice President of Applied Statistics and Data Science; and Dwarakh Nayam, Software Engineer, led the discussion. Students spent the morning learning about Reveal's Construction Re-Engineering project and the creativity, research, and analysis that is put into a federal data science initiative. 

April 2024

This month, Reveal presented four projects at two major industry convenings: the Federal Committee on Statistical Methodology and FedCASIC. At FCSM, Alex Zakrzeski showed off the ACS Autocoder, Pavani Samala talked about the implementation of a large language model for legislative research, and Jess Huang detailed efforts to improve response rates with better contact information. At FedCASIC, Taylor Wilson presented an overview of Census' AI initiatives to detail how we are bringing cutting edge LLM technology to federal statistics. 

March 2024

Reveal is proud to support the National Bureau of Economic Research (NBER) as a sponsor of the upcoming Conference on Research in Income and Wealth (CRIW) on Race, Ethnicity, and Economic Statistics in the 21st Century, taking place this spring. This conference will bring together leading scholars, policymakers, and data experts to explore critical issues surrounding race, ethnicity, and the evolving landscape of economic measurement. The CRIW Conference on Race, Ethnicity, and Economic Statistics in the 21st Century is an important forum for addressing challenges in data collection, representation, and analysis. As a sponsor, Reveal is excited to engage with experts in the field and contribute to a more comprehensive understanding of economics.

For more information about the conference, visit the conference website here.

November 2023

Reveal attended the William & Mary Tech Day Employer Showcase, hosted by the W&M School of Business, as an industry expert in generative AI. The theme of the day was "Age of Smart Technology & Radical Change."

October 2023

Reveal attended the Puerto Rico IT Cluster in San Juan to support the creation of solutions to help collect better data in Puerto Rico through address harmonization. The Puerto Rico IT Cluster is an alliance of industry, government, and academic partners committed to position Puerto Rico as a national and global leader in information technology services and innovation.

February 2022

Reveal is working with the Census Bureau to modernize its Construction Program by supporting the implementation of satellite imagery to track the progress of construction sites and starts. This modernized survey collection methods dating back to the 1950s. Read more here.

September 2021

This article explains "emotion AI," or E-AI, how it can be used by companies and organizations, and risks associated with it, concluding with a suggestion for readers to consider its potential limitations and consult with Reveal to develop a risk mitigation strategy. Read more here.

May 2021

Reveal Vice President for Research and Development Amish Chudasama outlines three steps business leaders can take to develop a sound AI governance stategy: build a compelling business case; come up with a strategy that guides development of AI embedded across an organization's ecosystem; and seek guidance from AI governance experts. Read more here.

April 2021

Reveal Vice President for Research and Development Amish Chudasama makes the case for business leaders to incorporate AI governance into their organizations, arguing that AI is increasingly being used by businesses and needs to be implemented in an understandable, transparent, and ethical manner. Read more here. 

March 2021

This article discusses AI's capability to finely analyze images in a variety of contexts, including healthcare, life science, and the US Food and Drug Administration (FDA). These capabilities allow for drawing insights more efficiently than if one were just using the human eye, and provide upskilling opportunities for workers. Read more here.

January 2021

This article explains the problem of fraud, waste, and abuse (FWA) in the healthcare sector, and the necessity of AI solutions to detect it given the large swath of records in the industry. Reveal offers its services in creating and implementing solutions that analyze large amounts of data to look for FWA and non-compliance. Read more here.