Separate the signal before the referral.
Bilingual students who do not exit ELL status by Grade 3 are 156% more likely to be misclassified. XODA provides the diagnostic layer that standard systems were not built to supply. (Cruz & Firestone, 2022)
The math is still there.
It was never missing.
The tool that could see it — that was missing.
Standard assessments run mathematical ability and language fluency through the same lens. For a student still developing English, that lens measures both at once — and reports one score.
When a bilingual student in a Title I classroom writes their math reasoning in abbreviated, phonetic, or mixed-language form, a standard rubric sees incomplete work. The diagnostic system built here sees a different signal: the mathematical structure beneath the surface.
That separation is the design. It is not an accommodation. It is the architecture.
Bilingual students who do not exit ELL status by Grade 3 are 156% more likely to be misclassified. XODA provides the diagnostic layer that standard systems were not built to supply. (Cruz & Firestone, 2022)
Fifteen years in Title I classrooms. The pattern was the same every time: the math was always there. The students who write in consonant skeletons, who mix Spanish and English mid-sentence, who abbreviate every word — they are not behind in mathematics. They are ahead in code-switching.
XODA reads the math signal the same way an experienced teacher does. Architecturally. Without guessing.
If your child solves problems in their head, explains solutions in a mix of languages, or writes answers that look abbreviated — those are not gaps. They are signals. XODA is built to read them.
QuizMyBrainz was founded by two educators who brought over 30 years of combined Title I bilingual classroom experience across multiple North Texas school districts to a single diagnostic question. Every misclassification they witnessed — every bilingual student routed toward an intervention that tested language fluency instead of mathematical ability — sharpened the same question: why doesn't a tool exist that separates these two signals?
The answer was that building it required both classroom-level ground truth and systems-level architecture. That combination took over 30 years to accumulate and one focused research sprint to deploy.
2015 Fort Worth ISD Campus Teacher of the Year, recognized by Texas House Resolution H.R. 2522, 84th Legislature.
Peer-reviewed manuscript under preparation for BEA 2027 (ACL Workshop on Innovative Use of NLP for Building Educational Applications).
Raw student text never reaches any AI model.
That is not a policy commitment.
That is the architecture.
FERPA and HIPAA Safe Harbor compliant.
How we protect student data →