Introducing XtraMath Pro with Fluency Intelligence

XtraMath Pro with Fluency Intelligence turns daily adaptive practice into weekly instructional clarity.

A student can understand the day’s lesson and still fall behind. The root cause might be invisible to the teacher, inadequate fact fluency.

The student may know what a fraction represents or understand how an equation works. But in complex or mult-step problems, calculation competes with reasoning for the same limited working memory. By the time the student reaches the new mathematics, much of that capacity has already been spent.

Teachers see the symptoms: slow work, unfinished problems, fragile confidence, avoidance, and growing anxiety. What is harder to see is the unfinished fluency underneath.

In “Agency over Agents,” we argued that the best role for AI in education is not to replace the teacher. We believe AI must reduce the assessment, alignment, and governance burden that limits educators ability to provide timely and coherent support for each student.

Today, we are introducing the next major step: XtraMath Pro with Fluency Intelligence.

The foundation matters

It's not sufficient just to memorize facts. Fact fluency does not guarantee mathematical proficiency, and it cannot explain every student's struggle.

But accurate, efficient, and flexible access to basic number relationships creates capacity for more advanced thinking. By middle school, students must coordinate fractions, ratios, rational numbers, expressions, equations, and multi-step problems. When basic calculation continues to demand significant attention, all of mathematics becomes harder to access.

The anxiety and struggle leads too many kids to identify as "not a math person.” Schools need a way to see this constraint earlier, before it appears only as a middle-school intervention problem. Ultimately this caps the number of students that will succeed in Algebra, STEM, or AI fields.

The challenge is not a lack of curriculum, materials, or data. It is the time and expertise needed to turn the evidence into actionable instructional clarity.

The signals are already there

Since 2009, tens of millions of students have used XtraMath’s short, adaptive practice to develop fact fluency.

That familiar classroom routine of 10-minutes a day does more than provide retrieval practice in each student’s zone of proximal development. Each session creates evidence about a student’s progress from effortful calculation toward more accurate, efficient, and flexible recall.

Students do not need another assessment event or a conversation with an AI tutor. Teachers do not need another dashboard filled with information they must analyze on their own.

The signals are already there. Fluency Intelligence transforms them into actionable instructional insights.

From daily practice to a weekly brief

Fluency Intelligence works in the background of XtraMath’s established practice routine. It assesses recent evidence, aligns supported patterns with research-based mathematical strategies, and audits each finding before it reaches the teacher.

The result is a concise Weekly Classroom Brief designed to help educators see where students may need attention and decide what to do next. The system performs the synthesis. The teacher brings the context, the relationship, and the instructional judgment.

This distinction matters. Practice evidence can reveal patterns, but it is not a diagnosis. It cannot capture a student’s complete mathematical understanding, determine potential, or explain everything happening in a classroom.

Fluency Intelligence is designed around those boundaries. When the evidence does not support a conclusion, the system should not manufacture one.

The AI harness for educators

Fluency Intelligence is not a synthetic teacher. It does not chat with students, replace the district’s curriculum, or make final instructional decisions.

It takes on work that educators could do with unlimited time: examining frequent practice evidence, identifying supported patterns, connecting those patterns to appropriate strategies, and checking the resulting guidance for unsupported claims.

That is what we mean by agency over agents.  Our platform is designed to use the minimum amount of screen-time needed to gather evidence necessary to make instructional decisions significantly more effective. It uses a curated knowledge graph with research-based best practices connected to high-quality instructional materials. 

We are building a governed, evidence-based platform specifically designed for math educators. In this content, we believe LLM-based Agents can be helpful, but must be integrated into the teacher's workflow, not get between the teacher and the student. It should make the education system more coherent without asking the teacher to learn or do something new.

Making the invisible actionable

Starting in August 2026, XtraMath Pro with Fluency Intelligence will be available for approximately $1-$2 per student.

The learning-science foundation behind this work is well established. The outcomes of this particular product must still be measured. We intend to approach that work with the same discipline we bring to the product itself: clear evidence, explicit boundaries, responsible governance, and respect for educator judgment.

Over time, we want to provide Fluency Intelligence as digitial public utility for PK-12 math education across the country and the world.

For teachers to district math leaders, the opportunity is straightforward: use the daily practice students already use to build a critical mathematical foundation earlier and get greater instructional clarity, coherence, and confidence.

Visit our Fluency Intelligence site to view a sample Weekly Classroom Brief and join the waitlist.

Roy King,
Executive Director

‍