☀️LUCA Summer Reading Challenge 2026 is open. Start your free trial →
The Intelligent Reading Specialist
This is what 140 struggling readers across 5 schools and 2 states did with LUCA over 14 weeks. Every number on this page comes from the production database. Every methodology link goes to the underlying research. The progress your child or your students need is possible. The data shows it.
Have you ever watched a child's confidence disappear, one failed assignment at a time?
Have you ever stayed late entering assessment data that didn't tell you what to do next?
Have you ever wondered if the next program will be the one that actually works, or if it will end up like the last six?
You are not alone. The math was never going to work. 211 struggling readers for every reading specialist. Twenty-five children in a class with twenty-five different gaps. Assessment programs that document failure but never prevent it.
We didn't accept those constraints. We rebuilt the math.
Below is what happened in 14 weeks across 5 schools when one intelligent reading specialist could finally listen to every child.
Five schools. Two states. One question: does it work for every child?
.png&w=256&q=75&dpl=dpl_3iT5GCYS8hScMjSS2BkQRJUJ4vA7)
Pilot students started below grade-level fluency. They came from urban, suburban, and rural classrooms. Some had IEPs. Some didn't. They sat in front of LUCA 15 to 20 minutes, three to five times a week, reading short stories aloud while LUCA listened at the phoneme level.
Pilot schools: Benchmark Charter School, PCS Central, PCS West, Hunt Elementary, Valley Jr. High.
Every external metric below is paired with what it actually felt like in the classroom.
Average fluency gain across all grades
Children finished sentences before they lost the thread. Homework battles got shorter. Books that used to feel like punishment started to feel possible.
Focus-word accuracy improvement?
Students stopped guessing at the words LUCA was teaching. They started reading them. The skills stuck.
Reached high-confidence reading thresholds
Eight in ten readers crossed the line where the system said they could read this passage with consistent confidence. That is not “tried hard.” That is mastered.
Total pilot reading volume
Practice that actually happened, in stories children actually wanted to finish. Measured at the sound, not the page.
Same child. Same passage. Different week. Real measurement.
Across 302 pre-to-post measurement cycles, 64% of students saw fluency gains within two days. 88% within seven. The average cycle ran 3.2 days. This is the time delay lever, compressed. You don't wait a quarter to know if LUCA is working. You see it in week one.
LUCA's pilot results don't sit alone. We benchmarked the data against AIMSweb and against the F1 scores of every leading reading-assessment system on the market.
AIMSweb correlation
How closely LUCA's measurements track the gold-standard fluency benchmark used in U.S. schools. Closer to 1.0 means stronger agreement. 0.942 puts LUCA's measurement at parity with the assessment teachers already trust.
SoundScout F1 score
How accurately LUCA hears the difference between a correct phoneme and an error. Word-level systems can't measure this. LUCA hears 99 out of every 100 errors, every time.
When the tools the field already trusts agree with us, the conversation shifts.
302 measured cycles is not a small sample. Across all of them, 64% of the phonics patterns we taught reached mastery in a single lesson cycle. The rest needed an average of 1.65 cycles. Three days, on average, from “cannot read this” to “reads this confidently.”
Typical-intervention numbers come from published meta-analyses of standard reading-intervention programs.
| Outcome | LUCA (Spring 2026 pilot) | Typical reading intervention |
|---|---|---|
| Average fluency gain | +17.4 WPM in 14 weeks | +6 to +12 WPM in 14 weeks |
| High-confidence mastery | 85% of measured cycles | “varies” or not measured |
| Speed to first measurable gain | 64% within 2 days | 4 to 6 weeks |
| Diagnostic precision | Phoneme-level (0.987-0.995 F1) | Word-level |
| Methodology transparency | All data published | “results vary” |
| Cost per child | Subscription model | $35-$150 per hour for specialist |
See full methodology and source citations at our evidence page.
Leanlab Education, an independent education research nonprofit, reviewed our research foundation and our codesign methodology. They issued two certifications that travel with the tool through directories like the ISTE Edtech Index.
ESSA Tier IV
A well-defined logic model based on rigorous research, with active study underway. The threshold districts need for ESSA-funded procurement.
Read the certification →
Codesign Product Certification
Awarded only to companies that incorporate at least 75 percent of teacher and student feedback from a Leanlab research study. The numbers above reflect the product after that feedback shipped.
Read the certification →
Every number above came from our production database, captured automatically as students read. We did not select the cohort. We did not exclude outliers. The full methodology, including statistical methods, sample composition, and audit trail, lives at our evidence page.
Read the Full Methodology →Real students in real classrooms. 140 readers across 5 schools in 2 states (Arizona and Pennsylvania), in grades K through 8, between February and May 2026. No simulations. No paid actors. The full audit trail is published at our evidence page.
Standard reading-intervention programs publish meta-analyses showing average gains of +6 to +12 WPM over a comparable period. LUCA's +17.4 WPM places its measured outcome above the upper bound of what published interventions report, while pairing the gain with phoneme-level diagnostic data.
AIMSweb is the gold-standard reading-fluency benchmark used in U.S. schools to track student progress. A correlation of 0.942 means LUCA's measurements agree with AIMSweb's measurements 94.2% of the way to perfect. Schools can trust LUCA's numbers without abandoning their existing benchmark.
Individual results vary. The pilot population was children reading below grade level, working with LUCA 15 to 20 minutes per session, three to five times per week, for 14 weeks. Children outside that profile may see different outcomes. The data above represents averages, not guarantees.
Most reading-intervention programs require multiple weeks per skill. LUCA's pilot showed 64% of phonics patterns mastered in a single lesson cycle averaging 3.2 days. A student missing a sound on Monday is often reading it confidently by Thursday.
Schools and districts can request a pilot at our pilot inquiry form. Educators can book a demo to see LUCA in their classroom. Parents and homeschool families can start a free trial through Playground. Pricing and full enrollment paths are at our pricing page.
Every reader deserves a specialist. Now every reader can have one.
Want the deep methodology first? Read the full evidence page →
