XtraMath Pro with Fluency Intelligence

An X-Ray for Math Brains.

What if you could see student’s working memory as it gets overloaded, and what next best specific action you could take to increase that capacity?

Since 2009, XtraMath has helped tens of millions of students build math fact fluency with 10 minutes a day of adaptive practice. XtraMath Pro with Fluency Intelligence transforms that data into instructional briefs that provide teachers more clarity, coherence and confidence. Fluency Intelligence realies on an audited, evidence-based knowledge graph, not LLM halucinations to automatically assess student progress and aligning it approved guidance resources.

Weekly Classroom Brief for Jefferson 5th Grade Math — showing this week's strategy-practice plan with student groupings.
Fig. 1 · A real brief from the beta with student names changed.
  • ESSA Tier IV Certified

    Proven outcomes with a research-based logic model

  • WCAG 2.1 Level AA

    Accessible for students and educators of all abilities

  • Student Privacy Pledge

    Compliant with FERPA, COPPA, CCPA, and GDPR

  • 501(c)(3) Nonprofit

    Mission-driven organization dedicated to math fluency

01 · The demand

The last two decades included record investment in curriculum, intervention, and technology.

The underlying constraint — the ability of teachers and students to make sense of everything — is rarely addressed.

02 · The hidden constraint

Working memory is a hidden bottleneck.

Overload produces the same symptom on both sides. Neither is a failure of effort. Both are what limited working memory does when it is asked to hold too much at once.

In the student

A student can understand the lesson and still run out of working memory before the work is done. When basic facts still require counting or reconstruction, calculation competes with reasoning for the same limited capacity. The demands on working memory grow quickly: fractions, ratios, expressions, equations, multi-step problems. This is not a narrow elementary skill. Unfinished fluency quietly caps how many students reach Algebra and STEM, while only the symptoms are visible:

  • Slow or incomplete multi-step work
  • A concept understood but not applied
  • Avoidance, shutdown, low mathematical confidence
  • Anxiety rising as tasks get more complex
In the planning

A teacher’s week runs ~54 hours,1 and fewer than half of those hours are spent teaching. About a dozen of the rest go to planning, preparation, and grading; the work of deciding what each student needs next. No one can hold thirty fluency journeys in their head: where each student is, which facts are secure, which strategy comes next, and where that strategy lives in the curriculum the school already trusts. Without coherent and timely information, planning defaults to the middle of the room, and the students at both edges wait.

1. Merrimack College Teacher Survey, administered by the EdWeek Research Center — 1,324 teachers, nationally representative, fielded January–February 2022.

03 · How Fluency Intelligence works

The classroom learning loop

Students keep the same ten-minute daily practice millions of classrooms already know. Fluency Intelligence works between practice and instruction — assessing the evidence, aligning research-based strategies, auditing every claim — and returns a brief that respects teacher judgment.

The classroom learning loop with Fluency Intelligence CLASSROOM LEARNING Students + Teacher Adaptive Practice 10 MIN / DAY FLUENCY INTELLIGENCE Assess Align Audit Instructional Brief 10 MIN / WEEK
Fig. 2 · The classroom learning loop — instruction leads to practice, practice creates evidence, and the brief returns to teacher judgment.

04 · The Instructional Brief

A jump-start to the week

One page, ten minutes, based on your students’ actual recent practice. Here’s a real one, section by section.

  1. Students are grouped by the evidence.

    Brief section: TEACH — Strategy practice, 10 students. Students share the same operation band and are working at the edge of the practice zone. Shows the Addition: Make 10 group — Viviana Sosa, Matias Eshetu, Taylor Miles — with the Making Ten strategy and a linked Illustrative Mathematics task.
  2. Each group gets a research-based strategy — linked to materials.

    Brief section: Subtraction — Doubles Reverse group for Gianna Martinez and Paislea Sanderson, with the Think Addition for Subtraction strategy and a linked Illustrative Mathematics task, "Hitting the Target Number."
  3. When the evidence is thin, the brief says so.

    Brief section: RECOVER — Evidence recovery, 2 students. Students have no recent sessions or are below the practice threshold. Brooklyn Sargent and Cali Quintero are flagged to get two short sessions before interpreting.
  4. Every student at a glance: mastery, change, and this week’s move.

    Brief section: Student briefs (20) table, sorted by attention needed, with columns for Student, Operation, Mastery, Change, and This Week’s Move — showing the first four rows: Dasha Garcia, Judith Cayetano, Raine Jones, and Justin Price.

Names are pseudonyms; the brief is otherwise exactly what teachers received.

05 · Governance

Governance by Design.

Every finding is audited before it reaches a teacher for unsupported claims, invented information, and deficit framing. When the evidence is insufficient, the brief says so. The teacher always makes the final decision.

  • No student chat

    Fluency Intelligence never talks to children. Analysis happens behind the teacher.

  • No automatic decisions

    Recommendations, never prescriptions. Placement and intervention stay with educators.

  • No extra screen time

    The same ten-minute routine students already know. No new apps, no new logins.

  • No teacher evaluation

    Briefs inform teachers. They are not surveillance and they don’t grade anyone.

Side by side

No chatbot. No black box.

Most AI products put a model between teacher and student. Fluency Intelligence puts AI to work for the teacher by assessing practice and aligning evidence in compliance with district policy.

AI-first products Fluency Intelligence
Generates more materials Teachers determine instruction
Students interact with AI Students engage with teacher and peers
More content to manage Content managed automatically
Fewer decisions to make Decisions backed by evidence
What's most profitable What's most important
More screen-time for students Encourages 10-minutes-a-day consistency
Requires prompting Performs recurring tasks
Adds workflow Compresses time and effort
Generic AI outputs Evidence-grounded interpretation
Replaces teacher judgment Respects teacher roles
  • Students do not need another app, another login, or a conversation with an AI tutor.
  • Teachers do not need another dashboard to learn or another place to check. They need the action to take, not more data to interpret.

From the classroom

“I have watched my students go from being frustrated and dreading math to being excited to log in to XtraMath!”

Brianna · Classroom teacher

The offer · 2026–27 school year

~$2 per student per year.

The daily practice routine students already know, plus a weekly instructional brief for every teacher. Standard XtraMath stays free — for any teacher, any family, forever.

FAQ

Questions district and school leaders ask

Is Fluency Intelligence a chatbot my students will talk to?
No. Students never interact with AI. Fluency Intelligence runs behind the teacher, assessing practice evidence after it’s created. The student experience is the same ten-minute routine XtraMath has always offered.
Does it add screen time?
No new apps, no new logins, no new student-facing software. Students keep the same ten-minute daily practice; teachers get a ten-minute weekly read. The point is to give time back to instruction.
Is XtraMath still free?
Yes — forever. Standard XtraMath stays free for any teacher, any family. Fluency Intelligence is part of XtraMath Pro, which costs ~$2 per student per year.
What data does it use, and is it private?
Only the practice evidence students already generate in XtraMath — what they answered, how fast, on which facts. XtraMath is FERPA and COPPA compliant, ad-free, and never sells data. We’ve been a 501(c)(3) nonprofit since 2009.
Does it evaluate teachers or place students?
No. Briefs are recommendations for professional judgment — placement and intervention decisions stay with educators. It is not surveillance and it doesn’t grade anyone.
How is this different from ChatGPT or an AI tutor?
Generic AI generates plausible text. Fluency Intelligence is a governed pipeline: it only reports what practice evidence supports, audits every claim before a teacher sees it, and withholds what fails the audit. The full method is in the white paper.