Using Predictive AI Models to Accelerate Growth and Close Achievement Gaps 

Reporter: Casey Welscher, University of Illinois Springfield 

Panelists: Jillian Tsoukalas, Ed.D., Assistant Superintendent, Troy CCSD 30C; John Bruesch, Ed.D., Vice President, ECRA Group, Inc.; Ryan Newberry, Director of Special Education, Troy CCSD 30C; Karina Ochoa, M.D., Director of Multilingual Services, Troy CCSD 30C 

Troy CCSD 30C shared how predictive analytics and AI-powered modeling are helping the district accelerate growth for all students while addressing persistent achievement gaps among English Learners and students receiving special education services. With an enrollment of approximately 3,900 students — 16% of whom qualify for special education and 4% as low income — the district is in its second year of a strategic plan centered on continuous improvement and data-informed decision-making. 

A key feature of the plan is the district’s “data scoreboard,” a framework that supports teams in writing targeted action plans and conducting regular impact studies. Presenters noted that new state benchmarks and percentile growth data prompted the district to refine its analysis, paying particular attention to subgroup performance. By examining state percentile trends, growth percentiles, and MAP Growth reading data, teams were able to identify where curriculum implementation was succeeding and where adjustments were needed. 

The district has prioritized deeper analysis of multilingual learners and students with Individualized Education Programs. A Special Education Leadership Committee meets regularly to review district-level and school-level charts, interpret assessment trends, and guide instructional decisions. Similarly, Multilingual Services staff work closely with data to understand student progress and identify supports. The district has also strengthened its parent engagement through multicultural nights and parent group meetings to help families understand assessment data and learning goals. 

ECRA Group’s predictive analytics tools, using 8–10 years of historical and local data, provide projections that help the district set realistic, achievable goals for students. Teachers participate in structured data reviews, where AI-informed projections and progress monitoring results from tools such as FastBridge are used to determine specific interventions. Presenters emphasized the importance of filtering data to identify students showing low or limited growth, allowing staff to respond with tailored support. 

Professional learning has been central to the district’s success. Specific areas targeted for development, including U-Fly training, have helped build educator confidence and consistency in using data across schools and teams. Presenters shared that scaling this training has resulted in more purposeful conversations about student progress and instructional strategies. 

The panelists concluded that predictive AI modeling has strengthened the district’s continuous improvement cycle by making data more actionable and by aligning interventions with specific student needs. For boards seeking to replicate this work, they highlighted the value of investing in high-quality data systems, building staff capacity, and ensuring that families are informed partners in the improvement process.