See who's struggling before the midterm says it out loud — learning analytics read by educators, sharpened by models that show their work.
Smart Edu AI is a Canadian edtech studio on West Broadway, Vancouver. We design and deliver AI-driven learning analytics, adaptive learning models and outcomes insight for schools, colleges and universities — with educators in the loop, student-data privacy protected, and academic integrity respected. We are not a credential seller, not a wellness or life-coaching brand, and not an "AI income" course.
NOTE · SE-01 · TERM-VIEW
A Broadway learning studio for evidence-minded educators
Dropout was predictable by week three — but nobody on faculty saw it until the withdrawal list landed. An accessibility gap surfaced only at exam time. A dashboard full of charts that no department chair could read in under twenty minutes. These are the problems we take on: learning analytics views, knowledge tracing models and outcomes insight reports that inform human decisions instead of replacing them.
Smart Edu AI is an education-AI consultancy serving client organizations across K-12, higher education and training providers in the Canadian market and wider North America. Senior learning scientists run discovery, build learner modelling profiles, and deliver production reporting with responsible-AI assurance, PIPEDA-aligned data governance and clear documentation at every stage.
DATA-TO-INSIGHT METHOD
Assess · Adapt · Teach · Measure
SE-01
Assess
Audit data quality, cohort definitions and what faculty actually need to decide — before any model is fit.
SE-02
Adapt
Personalised learning path recommendations with prediction confidence ranges, not false certainty.
SE-03
Teach
Outcomes insight reports and at-risk early signals placed in workflows instructors already use.
SE-04
Measure
Retention analysis reads and cohort performance trends tracked against agreed measurable outcomes.
RESPONSIBLE AI
Models are uncertain — we say so in the footnotes
AI in education fails when analytics pretend to be oracle cards. Our learner modelling profiles include confidence intervals, data-quality flags and educator sign-off gates. Knowledge tracing informs tutoring adjustments; it does not auto-expel anyone. Student-privacy safeguards and BC FIPPA / PIPEDA compliance are documented from discovery through production delivery.
Human-in-the-loop review is standard. Inclusive-learning features and accessibility audit passes run before any dashboard reaches a dean's inbox.
NOTE · SE-06 · CAPABILITIES
Six disciplines on the coral footnote rail
Learning Analytics View
Cohort performance trends faculty can act on this week.
Adaptive Learning Models
Personalised learning paths with documented limits.
Knowledge Tracing Model
Skill mastery estimates with prediction confidence ranges.
Assessment Support
Automated feedback drafts for educator review — never auto-issued.
Inclusive-Learning AI
Accessibility audit pass built into analytics pipelines.
Responsible-AI Assurance
Model QA, bias review and student-privacy safeguard documentation.
SELECTED WORK
Two anonymised engagements
Vancouver school district — at-risk early signal
Grade eleven math attrition followed a pattern visible in LMS clickstreams by week three. We delivered an at-risk early signal view with educator sign-off; counsellors contacted students before midterm. Illustrative past result: 14% fewer late withdrawals. Not a guarantee for your district.
Canadian university — outcomes insight report
Faculty senate wanted retention analysis reads tied to programme interventions, not vanity metrics. We built a first data cut to briefed insight pipeline with responsible-AI assurance. Past client work only.
MINI FAQ
Is .life a wellness brand?
No. The TLD is branding only. We are a learning-analytics edtech studio, not life-coaching or self-help.
Do you guarantee retention?
No. Models inform educator decisions; they do not guarantee grades, retention or employment outcomes.
Typical project budget?
Analytics reviews from C$14,000; full engagements C$40,000–C$110,000. Details in FAQ.
Request an analytics review
Bring your cohort question. We will tell you what the data can support — and what it cannot.
Request an analytics review