Building a tutoring nonprofit that now serves 25,000 students a year — with a vision to reach millions — required someone who'd lived the exact gap she was trying to close. Aly Murray grew up a low-income student raised by a single immigrant mother, moved through a string of Title 1 schools, and felt firsthand what it's like to navigate homework and college applications without support at home.
She left a trading job at JP Morgan eight years ago to build Upchieve, a nonprofit offering free, 24/7 human tutoring and college counseling to every Title 1 middle and high school student in the country.
Her work is backed by a Gates Foundation-funded study comparing human and AI tutoring, and Upchieve now partners with schools, districts, and CMOs at a cost of about 50 cents per tutoring session.
Most school tutoring programs run from 3 to 5pm — and that single design choice quietly locks out the students who need help the most. Aly Murray, founder of the nonprofit Upchieve, built a 24/7 human tutoring program instead, and the data on why AI can't replace it yet might surprise you.
📚 What You'll Learn
- Why the standard after-school tutoring program design quietly excludes the students who need it most
- What a Brookings Institute report reveals about grades, identity, and college enrollment
- Why Upchieve's founder built a 24/7 human tutoring model instead of scaling with AI
- What a Gates Foundation-funded study found when students had access to both AI and human tutors
- How much it actually costs a school to close its tutoring gap
💥 Breaking Down the Old Rules
🚀 Key Insight #1: Why Your After-School Tutoring Program Is Built to Exclude Kids
- What's broken: Drop-in, after-school tutoring assumes every student has a ride home, free time, and no caregiving responsibilities.
- The shift: Upchieve makes human tutoring available 24/7, matching a student to a live tutor in about two minutes — so support meets students wherever and whenever they actually need it, including 3am calculus.
- Impact: Upchieve gets over a third of students at partner schools using the platform regularly, a usage rate most after-school tutoring programs never come close to.
🚀 Key Insight #2: The Data Says Tutor the Middle, Not Just the Bottom
- What's broken: Most tutoring programs target only the lowest-performing 10-20% of students who are actively failing.
- The shift: Brookings Institute research found that once you control for academic preparation, over 70% of the college enrollment gap between low-income and high-income students disappears — meaning the biggest leverage sits with the "middle majority" of B and C students, not just the students at risk of failing.
- Impact: Small, consistent tutoring shifts a student's grades and, more importantly, their identity as "a student who's good at school" — which is what actually predicts whether they enroll in college.
🚀 Key Insight #3: Human Tutoring Still Beats AI Tutoring — And the Research Proves It
- What's broken: Schools are defaulting to AI tutoring because it looks cheaper and easier to scale than finding enough human tutors.
- The shift: In a Gates Foundation and Microsoft Research-funded study, less than 20% of students with access to both a human and an AI tutor ever tried the AI tutor once, and more than 92% of all tutoring sessions in the study were human-only.
- Impact: Students overwhelmingly chose — and stuck with — human tutors, because the motivational effect of a real person saying "I'm rooting for you" is what actually drives learning outcomes.
📣 ALY MURRAY QUOTES FROM THE RUCKUSCAST
"Tutoring is completely misunderstood a