Lirvana Labs, in partnership with District 10 of the NYC DOE, created a comprehensive learning ecosystem that prioritized both academic achievement and student engagement. Over the course of 5 weeks across 10 Summer Rising schools sites, more than 1,000 K-5 students actively participated in the program, spending an average of 19 minutes per day (75 minutes per week) on the Yeti platform. Leveraging Yeti Confetti Kids, one of Lirvana Labs’ student-facing products, and Yeti Co-Pilot, the teacher-supporting AI companion, these schools joined the Lirvana Labs Mountain Peaks Project to create an advanced AI model for elementary education.
Personalized assignments aligned to standards with support for 28+ languages and embedded Social-Emotional Learning (SEL) components
Real-time insights enabling teachers to track progress, identify gaps, and personalize instruction for all learners, including MLL and SPED students
The pilot program engaged students across 10 Summer Rising school sites throughout District 10, creating a comprehensive learning ecosystem that prioritized both academic achievement and student engagement.
This daily engagement was strategically reinforced through teacher-led activities that ensured seamless alignment with classroom goals and learning objectives. The program’s design recognized that meaningful learning happens when technology and human instruction work together harmoniously.
“I was able to see where the students levels were at, since they were not my students last year and I do not have any data of their understandings. Ask Yeti, I was able to modify the lessons for one student of my class that has an IEP and is below grade.”
- Kindergarten Teacher
63K
Total questions answered across ELA (54%) and Math (46%) subjects
79%
Exceptional performance across Math and ELA, with top classes achieving 92–94%
97K
Student achievements celebrating persistence and growth mindset
The Bronx pilot exceeded expectations across student learning and engagement. Students answered 63,707 assessment questions (an average of 60 each, ranging from 24394) and achieved a 79% accuracy rate across Math and ELA.
Perhaps most significantly, the program’s top-performing classes reached exceptional accuracy rates of 92-94% in a single week, demonstrating the power of personalized, adaptive learning technology. Teachers consistently reported that Yeti’s innovative design not only “explains the answer” to students but also helps them “stay engaged and motivated” throughout their learning journey. The multilingual support and differentiation capabilities proved particularly valuable for MLL and IEP students, creating truly inclusive learning environments. Over three weeks, our Educator Experience team visited all 10 Bronx schools in Summer Rising, helping teachers quickly implement Yeti Confetti Kids and Yeti Co-Pilot; onsite support ensured smooth adoption, boosted confidence, and proved our model’s scalability.
With the initial 10 schools successfully funded by MITSolve, Lirvana Labs has established a clear pathway forward for educational transformation in the Bronx, NYC, and beyond. The AI school house model, featuring Yeti Co-Pilot and Yeti Confetti Kids, represents a comprehensive approach that serves educators, students, and administrators equally. This fall, Lirvana Labs will continue supporting District 10′s growth with plans to expand to additional schools within the district and citywide. To support this growth, the Lirvana Labs Educator Experience team will launch a series of professional development for the fall and ongoing.
The vision extends far beyond a single district. Lirvana Labs is committed to expanding this proven model to partners throughout NYC and nationwide, creating a sustainable ecosystem that supports AI-powered learning for entire educational communities. This initiative represents more than just a pilot program—it lays the foundation for creating AI-powered districts that prioritize equity, engagement, and excellence in education for every student.
COPPA, FERPA, and ERMA Certified
© 2025 Lirvana Labs Inc. All rights reserved.