Twice a year, Education Analytics (EA) staff from across the country gather in Madison, Wisconsin, for collaborative meetings, team retreats, and skill-building. In July 2026, we hosted our second annual EA Conference, which included a full day of sessions to share the work, ideas, and expertise that define how EA operates and evolves.
Building on the energy of last year’s inaugural event, this year’s conference featured more than 25 sessions spanning five themes: Product, Partner, Platforms & Tools, AI, and Showcase. New this year was the Poster Showcase, a format that gave staff the chance to share projects in progress, spark conversations, and invite feedback in an informal, exploratory setting.
Taylor Odle, Ph.D., Assistant Professor of Educational Policy Studies at the University of Wisconsin–Madison, delivered the keynote, sharing his research on college access and success — including college admissions practices, financial aid, and advising and coaching — and what quantitative methods and data science can tell us about what keeps students engaged in postsecondary education. Odle’s work leverages field experiments with national partners and quasi-experimental evaluations of existing policies to study “what works” in states, systems, and institutions for improving students' transitions to and through college.
Several sessions explored how EA’s product suite is growing in depth and integration, and what that means for the partners who rely on it every day.
Featured Session | More Assessments, More Insight: How Cross-Team Coordination Drives Value to Partners
Presented by Julianna Alvord, John Raub, and Josh Terchek
This session explored how the Product Division and South Carolina Partner Pod expanded formative assessment loading through Runway — bringing richer, more current assessment signals into Rally and reducing the manual work required to get there. Presenters walked through the original bottleneck, the improvements that now allow administrators to load and reconcile assessment data more directly, and the downstream impact for partners.
Featured Session | Empowering Data Exploration: Enabling Users to Build Their Own Analyses
Presented by Molly Stewart, Rob Little, and Carly Benson
How do we support users as they move from pre-built analyses to creating their own? This session shared lessons learned from training users on Stadium and Podium+ — including two case studies from WInsights, South Carolina, and a South Carolina pilot for self-serve AI features in Stadium.
Featured Session | The EA Platform: Unifying Data to Drive Student Success
Presented by Rosh Dhanawade, Katie O'Brien, Molly Stewart, Joshua Terchek, and Silvia Brunet-Jones
This session walked through EA’s full product portfolio end to end — from StartingBlocks and Runway through Stadium, Podium, Rally, and the App Launcher Suite — tracing the data’s journey from intake to insight.
Two sessions highlighted the relational and operational work of supporting EA’s district and state partners across time.
Featured Session | Rolling in the Deep: Mastering a Partner Rollover
Presented by LaVetra Stevenson, Swathi Siva, and Meg McCarter
A Partner Rollover — the coordinated process of transitioning from one school year to the next across systems, data, vendors, and operational teams — requires months of careful planning and collaboration. This session explored the key phases of a successful rollover and gave participants practical strategies for managing two school years at once with confidence.
Featured Session | Can We Be Friends? Lessons from Communicating with 80 Districts in SC
Presented by Aldo Belmont, LaVetra Stevenson, and Brianna Pollari
After more than two years of weekly office hours and newsletters, EA’s South Carolina team developed a communication approach that keeps districts and internal teams genuinely in sync. This session shared how to gather useful feedback from partners and how to communicate in ways that build trust across a large, distributed group.
Featured Session | Metadata That Matters: Building the Foundation for Workforce Credentials
Presented by Jordan Mader, Christian Faulkner, Julianna Alvord, Sam LeBlanc, Drew McDermott, and Caroline Wang
What does it take to connect a student’s K–12 studies to the careers, skills, and training demanded by the modern workforce? This session used the Ed-Tech Collaboratory’s (ETC) Workforce Credentials project as a vehicle to explore how intentional, standardized metadata organization is critical to serving the public good.
A full block of spotlight talks dug into the systems and infrastructure that support EA’s engineering, delivery, and operational work.
Spotlight Talk | Heimdall: The Philosophy and Technology of Data Access for EA Products
Presented by Sam LeBlanc and Erik Joranlien
Heimdall determines which student data a district administrator, school principal, or teacher should see when they log into Podium or Rally. This session explained the ideas and technology behind how EA calculates and exposes security contexts for its external user-facing applications.
Spotlight Talk | Introducing Fauna: Building a Shared Synthetic Environment
Presented by Zhyldyz Abdykadyrova and Jacob Bortell
Fauna is an initiative to create a shared synthetic environment — one where engineers can test code, delivery teams can demo products, and end users can train on released features, all without touching personally identifiable information (PII).
Spotlight Talk | To Foundry and Beyond
Presented by Betsy Conti and Gabe Rosen
This session explored how teams can collaborate across functions to build productivity apps in Foundry and explore cross-team content for EA’s operational work.
Spotlight Talk | Connected by Design: Interoperability in EA’s Internal Operational Systems
Presented by Mark Stokosa, Ernie Morgan, Eric Cramer, and Quinn Beirne
This session showed how EA automated its Data Release Agreement (DRA) workflow by connecting commonly used applications — replacing manual tracking and siloed handoffs with alerts and integration into Foundry.
Artificial Intelligence was one of the most active threads of the day — appearing across featured sessions, spotlight talks, and the poster showcase alike.
Featured Session | Setting Your AI Up for Success
Presented by Craig Thompson and Greg Fitzgerald
The difference between “AI is amazing” and “AI is useless” is often the engineering around it. This session shared lessons learned from incorporating AI into real engineering workflows, along with a framework for thinking about context as a first-class engineering asset.
Featured Session | Vibe Check: A Vibe-Coded Talk About Staying Safe and Sane with AI
Presented by Tom Reitz and Mark TenHoor
A celebration of what AI makes possible — and an honest look at what happens when we treat it as a shortcut instead of a powerful tool. Using real-world examples explained in plain language, this session showed how trusted tools can quietly become vulnerabilities and made the case that if you’re building with AI, you need the people who can help build it, review it, and remember it exists.
Spotlight Talk | Leveraging Claude Code to Expand the Scope and Impact of Public Data
Presented by Elisabeth Nesmith
During EA’s org-wide AI Skunkworks venture, analysts explored how Claude Code could reduce the level of manual, time-intensive work required for the Public Data Project. This session gave a brief overview of the public data processes, highlighted the workflows piloted using Claude, and demonstrated how AI could transform the potential of public data into a valuable reality for both partners and internal quality control.
Spotlight Talk | Evaluating Machine Learning Approaches for Rally Calibration Models
Presented by Sara Hu
Rally calibration models predict student performance across assessments using historical data. This session compared a random forest model with simple and enhanced regression models across 1,752 calibration models — finding that random forests substantially outperformed simple linear regression, but their advantage disappeared against an enhanced regression model that captures key nonlinear relationships while offering greater transparency and interpretability.
Spotlight Talk | Forecasting? Rally Models? What Are These?
Presented by Hayley Tymeson
A plain-language walkthrough of EA’s forecasting model — the methodology, its history, recent updates, and model evaluation results. The session also previewed hopes for a future version of the forecasting model and connected directly to Sara Hu’s presentation on calibration model approaches.
For the first time, this year’s conference included a Poster Showcase, where staff shared projects in progress, invited feedback, and engaged colleagues in informal, small-group conversations. The format proved a natural fit for work that benefits from dialogue as much as presentation.
Showcase | What COVID Taught Us About Students’ Social-Emotional Development — and Why California Should Rethink How It Provides Support
Presented by Caroline Wang
Based on a recent commentary published with Policy Analysis for California Education (PACE) at Stanford University, this poster drew on longitudinal evidence from the CORE Districts to argue that California’s postpandemic recovery strategy should treat social-emotional development as a core dimension of student learning. Key findings: social-emotional competencies develop in nonlinear ways, were disrupted by COVID-19, and are associated with academic resilience — especially during key transitions such as the move into middle school and among students with disabilities.
Showcase | To Whom Are We Speaking? Knowing EA's Audience
Presented by Marlena Holden, Grayce Payne, and Evan Wooldridge
The Brand & Communications team shared EA’s main external audience profiles, the frameworks that guide how the team routes different types of communications, and the personas behind each profile — offering a practical reference for anyone at EA who creates, reviews, or contributes to external-facing content.
Showcase | Stadium AI Demo: Configurable Agents for Querying Data
Presented by Rob Little, Kayla Bollinger, Theo Kaufman, Drew McDermott, and Molly Stewart
The Stadium AI Working Group gave an interactive, behind-the-scenes look at the automated, configurable interface they’ve built for Stadium implementations to create AI agents — built to query, document, and analyze the data warehouse.
Showcase | Using Data Infrastructure to Know the Whys, Whens, and What-Matters of Student Absenteeism
Presented by Sara Hu
This poster shared research into how the impacts of student absenteeism on academic outcomes vary by reason for absence and by the extent of peer absence — using a novel data acquisition strategy that accesses large-scale live data streams through a standards-based, statewide data system.
Showcase | Evaluating Synthetic Data for Education Research
Presented by Andrew Rice and Rinor Jahiu
Can machine-generated student data serve as a reliable stand-in for real student data in education analytics? This poster presented findings from EA’s synthetic data research project, which trained and rigorously evaluated four generative models on a dataset of roughly 250,000 South Carolina students — and explored where synthetic data shows real promise for expanding EA’s offerings.
Showcase | ETC Assessment Integration Playbook
Presented by Julianna Alvord, Theo Kaufman, and Silvia Brunet-Jones
Assessment integrations often remain fragmented even when states and districts have adopted Ed-Fi. This poster introduced the ETC Assessment Integration Playbook, a practical framework for helping vendors build native, certification-ready integrations that support scalable, sustainable education data systems across the data ecosystem.
Showcase | What Do You Need to Succeed at EA? Help Shape EA's Employee Training Program
Presented by Marlena Holden and Kerrie Lentz
Human Resources and the Brand & Communication teams partnered on a real-time feedback session to help determine which people-centered training topics — from performance reviews to giving and receiving feedback — staff most want to explore.
Showcase | What Strings Us Together? A Data-Driven Creative Collaboration
Presented by Emily Arneson, Jessie Gallagher, Anna Geisthardt, Annie Kime, and Nick Schauer
Attendees built a large-scale parallel coordinates plot — one string at a time. Each participant answered a series of questions by weaving a single piece of string through a set of pegs representing their responses. The result: a collaborative work of art and a visualization of EA’s community.
Looking Ahead
We know that what makes EA strong is its people. The value of bringing staff together twice a year cannot be quantified, and staff leave our time together energized, inspired, and excited to channel their learnings to advance our work.
