At FDR High School in New York, an AI math coaching platform is being used for in-class support and homework—aiming to reduce math anxiety and strengthen skills aligned to NYC’s math curriculum push.
Math is supposed to build step by step, yet many students experience it as a moving target—especially when foundational gaps show up later.
Why math support is getting an AI upgrade
The challenge isn’t only student confidence.. Teachers also face the daily reality of mixed readiness levels in one room. limited time for individualized practice. and the pressure of keeping students progressing through the same grade-level curriculum.. Misryoum’s education beat sees a consistent thread in these efforts: schools want early intervention. interactive learning. and digital tools that can respond immediately—without waiting for a teacher to notice every misunderstanding.
Alignment with a curriculum schools can actually scale
That alignment matters because it reduces a common failure point in edtech rollouts: when a tool teaches skills that don’t match what students are learning in class. it can feel disconnected.. Instead. FDR teachers can reinforce concepts with the same lessons and activities embedded in the system. helping students practice the ideas they’re actually studying.
Misryoum also notes that FDR started using the AI platform for teachers in September 2024 before expanding to student-facing use, after data protection measures were in place. Now, depending on the class, the platform is used multiple days per week for targeted math support or as a homework tool.
Inside the coaching loop: feedback. retries. and progress signals
In one approach Misryoum observed, teachers use the platform primarily for homework.. A math teacher at FDR described how this supports a constructivist style of learning: classroom time can focus on building conceptual understanding. while the platform helps strengthen procedural fluency when students are working independently.. The teacher’s point is a balancing act that many educators recognize—conceptual learning can be slower. but digital practice at home can help students develop fluency without consuming every minute of limited class time.
There’s also a feedback-and-visibility element.. Teachers can see which questions students ask the AI coach. how many attempts a student makes on a problem. and what they repeatedly struggle with.. For educators managing multiple learners, that’s more than convenience; it can act like a window into student thinking.. Misryoum views this as part of a broader trend in education technology: shifting tools from being “answer engines” to being diagnostic engines that help teachers identify where misconceptions emerge.
What this could mean for students who need help most
FDR also has a large population of English Language Learners. and Misryoum reports that language translation features are a key part of why the tool is used.. Students can engage with activities while using their home language, which can lower barriers during early stages of English acquisition.. That’s not a small detail; in many math classrooms. language is often the hidden obstacle behind word problems and multi-step reasoning.
Teachers at FDR also describe an additional benefit: the platform supports self-reflection.. If students can track the kinds of questions they ask and where they repeatedly get stuck. they can learn to identify their own weak points and ask better. more targeted questions.. In Misryoum’s editorial framing. this is a shift from “help me” to “help me understand. ” which is crucial for independent learning.
Results and what to watch next
These numbers don’t automatically translate into passing rates. Misryoum cautions. but they do signal growth—something educators often look for when closing gaps.. Just as important. the model of use is specific: students are required to answer correctly before moving on. teachers have visibility into the learning process. and the platform is positioned as both a support system and a practice structure.
Looking ahead. Misryoum expects schools adopting AI coaching to keep refining two questions: how to maintain curriculum alignment at scale. and how to ensure students benefit from feedback loops without becoming dependent on the tool.. The most promising systems will be the ones that strengthen self-efficacy. improve misconceptions through retries. and help teachers intervene sooner—turning assistance into mastery rather than temporary relief.